PO-084 Familial risk of colorectal cancer in half-siblings similar to that in siblings

• Published in: ESMO open 2018-07, Vol.3 (Suppl 2), p.A259-A259
• Main Authors: Tian, Y, Kharazmi, E, Sundquist, K, Sundquist, J, Fallah, M
• Format: Article
• Language: English
• ISSN: 2059-7029; 2059-7029
• DOI: 10.1136/esmoopen-2018-EACR25.612

IntroductionFor improving evidence-based strategies for colorectal cancer (CRC) screening and making more informed decision about early counselling of CRC, we aimed to explore the risk of CRC in family members of CRC patients.Material and methodsUsing the world’s largest population-based Family-Cancer data from Sweden (follow up: 1958–2015), we calculated the lifetime cumulative risk (0–79 years) and relative risk of CRC for first- and second-degree relatives (FDRs and SDRs) of patients with CRC. All family histories reported below are exclusive, meaning that risk reported for one affected half-sibling does not include those with both an affected half-sibling and any other first/second-degree relatives.Results and discussionsA total of 15,880,734 persons were followed for up to 58 years, out of them 1 58 428 with clear and complete family information developed CRC. The overall lifetime risk of CRC in siblings of CRC patients was 7% (8% in men; 6% in women), which represents 1.7-fold (95% CI: 1.6–1.7; n=2055) increase over the 4% risk in those without any family history of CRC (5% in men; 3% in women). Significantly increased lifetime risk (6%) was found among half-siblings of CRC patients (SIR=1.6, 95% CI: 1.3–1.8; n=140), 8% in male half-siblings and 5% in female half-siblings. The risk in those with CRC in both a parent and a half-sibling (SIR=3.6, 95% CI: 2.5–5.1; n=32) was close to risk of those with an affected parent and an affected sibling (SIR=2.7, 95% CI: 2.4–3.0; n=391). Highly increased risk of CRC was also found in those with two affected half-siblings (SIR=3.5, 95% CI: 1.3–7.6; n=6). Family history of CRC in other SDRs such as a grandparent (SIR=1.2, 95% CI: 1.1–1.3; n=460), an uncle/aunt (SIR=1.2, 95% CI: 1.0–1.4; n=189) without any affected FDRs showed minor contribution to the familial risk of CRC, but we found higher risks for those with both an affected FDR and a grandparent (SIR=3.0, 95% CI: 2.4–3.8; n=82) or both an affected FDR and an uncle/aunt (SIR=2.2, 95% CI: 1.5–3.2; n=28), which approximate to that with two affected FDRs (SIR=2.5, 95% CI: 2.3–2.7; n=946).ConclusionFamily history of CRC in half-siblings contributes similar CRC risk to that in siblings. The CRC risk among those with one affected SDR was negligible except among half-siblings, but combination of family history in one affected SDR and an affected FDR (or even another SDR) increased the risk substantially. These evidence-based findings provide novel information to impr