Non-steroidal Anti-inflammatory Drugs, Polygenic Risk Score and Colorectal Cancer Risk

Xuechen Chen; Feng Guo; Michael Hoffmeister; Jenny Chang-Claude; Hermann Brenner


Aliment Pharmacol Ther. 2021;54(2):167-175. 

In This Article

Abstract and Introduction


Background: The regular use of non-steroidal anti-inflammatory drugs (NSAIDs) has been associated with reduced colorectal cancer (CRC) risk.

Aim: To explore whether this association varies according to background polygenic risk for CRC.

Methods: Data were collected from a large population-based case-control study on CRC in Germany. A polygenic risk score (PRS) based on 140 CRC-related risk loci was used to quantify the genetic risk. The associations of regular use of NSAIDs (≥2times per week for at least 1 year) with CRC risk were estimated in the whole population and in subgroups according to PRS levels using multivariable logistic regression. The impact of NSAIDs on CRC risk was compared to PRS using the 'genetic risk equivalent' (GRE), a recently developed metric for effective risk communication.

Results: In total 5129 CRC cases and 4093 controls were included in this analysis. The regular use of NSAIDs (including aspirin) was associated with reduced CRC risk [odds ratio (OR) 0.66, 95% confidence interval (CI) 0.59, 0.74], as was regular use of aspirin only (OR 0.73, 95% CI 0.65, 0.83), without indication of interaction with the PRS (P = 0.10 and 0.22 respectively). The effect of NSAID use was equivalent to the effect of having a 32 percentiles lower PRS (GRE −32, 95% CI −41, −22).

Conclusions: The regular use of NSAIDs is associated with greatly reduced CRC risk regardless of individual genetic profile. With an equivalent reduction of relative risk across all polygenic risk groups, absolute risk reduction would be expected to be strongest among those with the highest polygenic risk score.


Colorectal cancer (CRC) is a leading cause of cancer death worldwide, accounting for 9.2% of the total cancer deaths in 2018.[1] Colonoscopy screening with resection of adenomatous polyps has been associated with a great reduction in the incidence of and mortality from CRC but is challenged by low adherence in several countries.[2–5] Targeted chemoprevention strategies may have the potential to further reduce CRC incidence and mortality.[6]

Substantial evidence has demonstrated that regular use of non-steroidal anti-inflammatory drugs (NSAIDs), especially aspirin, is associated with reduced risk of colorectal adenomas and cancer.[7–12] However, in the light of potential adverse drug reactions such as the increased risk of gastrointestinal bleeding/ulcer and cardiovascular events, it is particularly important to identify those for whom the benefit outweighs the risk.[13,14] Inter-individual variation in the chemopreventive effect of NSAIDs on CRC has been in part attributed to germline variations, particularly single nucleotide polymorphisms (SNPs).[15–17] Recent studies have shown that incorporation of a polygenic risk score (PRS), aggregating multiple CRC-related SNPs identified in genome-wide association studies (GWAS), into traditional models could improve CRC risk stratification.[18,19] However, evidence on the interaction of PRS with NSAID use and the potential role of PRS in personalised use of NSAIDs for CRC prevention is lacking.

Thus, the aim of this study was to assess the effect of regular use of aspirin and NSAIDs (including aspirin) on CRC risk at different levels of PRS. We also compared the effects of aspirin and NSAID use with the impact of PRS on CRC risk using the 'genetic risk equivalents (GRE)', a recently developed metric for effective risk communication.[20]