Cause-specific mortality and multimodal risk determinants among 5-year ovarian cancer survivors

Abstract Background Evidence on causes of death beyond five years after ovarian cancer diagnosis remains limited. We characterized post-landmark cause-specific mortality and evaluated clinical, genomic, biochemical, and socioeconomic determinants among five-year survivors. Methods This cohort study used linked UK Biobank data from women who survived at least five years after ovarian cancer diagnosis. Follow-up was anchored at the fifth anniversary of diagnosis, with delayed entry at the later of this landmark and UK Biobank enrollment. Cause-specific cumulative incidence was estimated using the Aalen–Johansen method. Fine–Gray and cause-specific Cox models evaluated mortality from ovarian cancer, cardiovascular disease, second primary malignancies, and other non-cancer causes. Germline homologous recombination repair (HRR) variants were assessed in participants with whole-exome sequencing data. Results The cohort included 920 survivors, of whom 845 had sequencing data. During post-entry follow-up, 322 deaths occurred: 183 from ovarian cancer, 58 from cardiovascular disease, 42 from second primary malignancies, and 39 from other non-cancer causes. At five years after the survivorship landmark, cumulative incidence was 15.6% (95% confidence interval [CI], 13.3–18.1%) for ovarian cancer death and 7.5% (95% CI, 6.0–9.3%) for the combined competing causes. The adjusted risk difference was − 8.1% points (95% CI, − 10.4 to − 5.8), indicating persistent predominance of ovarian cancer mortality. However, differences were attenuated and CIs included zero among survivors aged at least 65 years, those with more than two comorbidities, and those in the highest deprivation tertile. Advanced stage and serous histology were associated mainly with ovarian cancer mortality. Older age, greater deprivation, higher comorbidity burden, adiposity, and current smoking were associated primarily with cardiovascular mortality; more than two comorbidities yielded a subdistribution hazard ratio of 3.32 (95% CI, 1.86–5.92). HRR carriers had higher second-primary-malignancy mortality than non-carriers (subdistribution hazard ratio, 3.74; 95% CI, 1.88–7.42), although this estimate was based on few events. Conclusions Ovarian cancer remained the principal late cause of death among five-year survivors, while competing mortality became increasingly important in older, comorbid, and socioeconomically deprived women.