Many individuals with multiple or large colorectal adenomas, or early-onset colorectal cancer (CRC), have no detectable germline mutations in the known cancer predisposition genes. Using whole-genome sequencing, supplemented by linkage and association analysis, we identified specific heterozygous POLE or POLD1 germline variants in several multiple adenoma and/or CRC cases, but in no controls. The susceptibility variants appear to have high penetrance. POLD1 is also associated with endometrial cancer predisposition. The mutations map to equivalent sites in the proof-reading (exonuclease) domain of DNA polymerases ε and δ, and are predicted to impair correction of mispaired bases inserted during DNA replication. In agreement with this prediction, mutation carriers’ tumours were microsatellite-stable, but tended to acquire base substitution mutations, as confirmed by yeast functional assays. Further analysis of published data showed that the recently-described group of hypermutant, microsatellite-stable CRCs is likely to be caused by somatic POLE exonuclease domain mutations.

Cancer chromosomal instability (CIN) results in an elevated rate of change of chromosome number and structure and generates intratumour heterogeneity1,2. CIN is observed in the majority of solid tumours and is associated with both poor prognosis and drug resistance3,4. Therefore, understanding a mechanistic basis for CIN is paramount. Here we find evidence for impaired replication fork progression and elevated DNA replication stress in CIN+ colorectal cancer (CRC) cells relative to CIN− CRC cells, with structural chromosome abnormalities precipitating chromosome missegregation in mitosis. We identify three novel CIN-suppressor genes (PIGN (MCD4), RKHD2 (MEX3C) and ZNF516 (KIAA0222)) encoded on chromosome 18q, which is subject to frequent copy number loss in CIN+ CRC. 18q loss was temporally associated with aneuploidy onset at the adenoma-carcinoma transition. CIN-suppressor gene silencing leads to DNA replication stress, structural chromosome abnormalities and chromosome missegregation. Supplementing cells with nucleosides, to alleviate replication-associated damage5, reduces the frequency of chromosome segregation errors following CIN-suppressor gene silencing and attenuates segregation errors and DNA damage in CIN+ cells. These data implicate a central role for replication stress in the generation of structural and numerical CIN, which may inform new therapeutic approaches to limit intratumour heterogeneity.

To identify colorectal cancer (CRC) susceptibility alleles, we conducted a genome-wide association study. In phase 1, we genotyped 550,163 tagSNPs in 940 familial colorectal tumor cases (627 CRC, 313 high-risk adenoma) and 965 controls. In phase 2, we genotyped 42,708 selected SNPs in 2,873 CRC cases and 2,871 controls. In phase 3, we evaluated 11 SNPs showing association at P < 10(-4) in a joint analysis of phases 1 and 2 in 4,287 CRC cases and 3,743 controls. Two SNPs were taken forward to phase 4 genotyping (10,731 CRC cases and 10,961 controls from eight centers). In addition to the previously reported 8q24, 15q13 and 18q21 CRC risk loci, we identified two previously unreported associations: rs10795668, located at 10p14 (P = 2.5 x 10(-13) overall; P = 6.9 x 10(-12) replication), and rs16892766, at 8q23.3 (P = 3.3 x 10(-18) overall; P = 9.6 x 10(-17) replication), which tags a plausible causative gene, EIF3H. These data provide further evidence for the 'common-disease common-variant' model of CRC predisposition.

To identify risk variants for colorectal cancer (CRC), we conducted a genome-wide association study, genotyping 550,163 tag SNPs in 940 individuals with familial colorectal tumor (627 CRC, 313 advanced adenomas) and 965 controls. We evaluated selected SNPs in three replication sample sets (7,473 cases, 5,984 controls) and identified three SNPs in SMAD7 (involved in TGF-beta and Wnt signaling) associated with CRC. Across the four sample sets, the association between rs4939827 and CRC was highly statistically significant (P(trend) = 1.0 x 10(-12)).

Genome-wide association (GWA) studies have thus far identified 10 loci at which common variants influence the risk of developing colorectal cancer (CRC). To enhance power to identify additional loci, we conducted a meta-analysis of three GWA studies from the UK totalling 3,334 cases and 4,628 controls, followed by multiple validation analyses, involving a total of 18,095 CRC cases and 20,197 controls. We identified new associations at 4 CRC risk loci: 1q41 (rs6691170, OR=1.06, P=9.55x10-10; rs6687758, OR=1.09, P=2.27x10-9); 3q26.2 (rs10936599, OR=0.93, P=3.39x10-8); 12q13.13 (rs11169552, OR=0.92, P=1.89x10-10; rs7136702, OR=1.06, P=4.02=x10-8); and 20q13.33 (rs4925386, OR=0.93, P=1.89x10-10). As well as identifying multiple new CRC risk loci this analysis provides evidence that additional CRC-associated variants of similar effect size remain to be discovered.

Accurate duplication of DNA prior to cell division is essential to suppress mutagenesis and tumour development. The high fidelity of eukaryotic DNA replication is due to a combination of accurate incorporation of nucleotides into the nascent DNA strand by DNA polymerases, the recognition and removal of mispaired nucleotides (proofreading) by the exonuclease activity of DNA polymerases δ and ɛ, and post-replication surveillance and repair of newly synthesized DNA by the mismatch repair (MMR) apparatus. While the contribution of defective MMR to neoplasia is well recognized, evidence that faulty DNA polymerase activity is important in cancer development has been limited. We have recently shown that germline POLE and POLD1 exonuclease domain mutations (EDMs) predispose to colorectal cancer (CRC) and, in the latter case, to endometrial cancer (EC). Somatic POLE mutations also occur in 5–10% of sporadic CRCs and underlie a hypermutator, microsatellite-stable molecular phenotype. We hypothesized that sporadic ECs might also acquire somatic POLE and/or POLD1 mutations. Here, we have found that missense POLE EDMs with good evidence of pathogenic effects are present in 7% of a set of 173 endometrial cancers, although POLD1 EDMs are uncommon. The POLE mutations localized to highly conserved residues and were strongly predicted to affect proofreading. Consistent with this, POLE-mutant tumours were hypermutated, with a high frequency of base substitutions, and an especially large relative excess of G:C>T:A transversions. All POLE EDM tumours were microsatellite stable, suggesting that defects in either DNA proofreading or MMR provide alternative mechanisms to achieve genomic instability and tumourigenesis.

This study confirms the intra-tumor stroma ratio as a prognostic factor. This parameter could be a valuable and low cost addition to the TNM status and next to current high-risk parameters such as microsatellite instability status used in routine pathology reporting. When adding the stroma-parameter to the ASCO criteria, the rate of 'undertreated' patients dropped from 5.9% to 4.3%, the 'overtreated' increased with 6.8% but the correctly classified increased with an additional 14%.

Implications of all the available evidence It is possible to utilise deep learning to develop biomarkers for automatic prediction of patient outcome directly from conventional histopathology images. In colorectal cancer, the marker was found to be a clinically useful prognostic marker in analysis of a large series of patients who received consistent, modern cancer treatment.