April 2025
·
10 Reads
Objective We aimed to analyze management and survival outcomes of older people (≥75 years) with stage II+ epithelial ovarian cancer (EOC) across six gynaecological cancer centers in the United Kingdom. Methods Retrospective cohort study performed using the IMPRESS project dataset. Clinical information for patients diagnosed with EOC from six sites of varying size and population demographics, was collated between 1/1/2018 and 31/12/2019. We compared treatment of patients aged ≥75 years with those <75, within and between centers, using multi-variate analysis, to understand effects on outcomes. Results After exclusions, we assessed 721 people for overall survival (OS) and 702 for progression-free survival (PFS). The ≥75s had poorer performance status, more comorbidities and were less likely to receive combination treatment with surgery and chemotherapy (in either order) (overall = 392/721 (54.4%) overall; <75 cohort = 320/495 (64.6%); ≥75 cohort = 72/226 (31.9%). Treatment proportions varied between sites, with some having no active treatment rates of nearly 50% for ≥75s. Those ≥75 years had twice the relative risk of death compared to those <75 (Relative Risk (RR) = 1.98, 95% confidence intervals (CI) 1.63 to 2.39, P<0.001). Adjustment for confounders individually caused only a relatively modest reduction in magnitude and strength of association. Adjustment for treatment led to this association essentially disappearing (RR = 1.10, 95% CI 0.88 to 1.38; 99% reduction in Chi2), though with significant variation in association between age and OS between treatment groups (p-heterogeneity: 0.0004). Conclusion Older women may do as well as younger women in terms of survival, if treated similarly, although this varies depending on treatment groups. Treatments varied between and within sites, with some sites treating older women more differently than others. Some differences may be appropriate, but significant differences in rates of no active treatment between sites suggests that not all variation may be appropriate.
