Background
There were relatively few substantive changes in ovarian cancer treatment in recent decades, until the introduction of bevacizumab in the early 2010s, and with the subsequent introduction of PARP (poly[ADP]-ribose polymerase) (PARP) inhibitors for subsets of cases, added to platinum-based chemotherapies. Improvements in outcomes of cytoreductive surgery occurred in parallel in many regions. 

Objectives
Population-based studies evaluating survival trends in the context of modern treatment are sparse. Our primary objective was to characterize ovarian cancer by histotype and stage, and by year of diagnosis, to evaluate trends in survival across grouped calendar periods during the time when targeted therapies were introduced. 

Methods
Nationwide data from the Center for Cancer Registry Data at the Robert Koch Institute and the Federal Statistical Office of Germany were used. Cases were diagnosed between 2010-2023. 

We defined the following four calendar periods of diagnosis: 2010–2012 (largely pre-targeted therapy period), 2013-2016, 2017-2019, and 2020-2023 (targeted treatment period). Excess hazard and relative survival were estimated using flexible parametric models within a relative survival framework. Expected mortality rates were derived from German life tables, stratified by age, calendar year, and federal state. To predict excess hazard for each selected calendar period of diagnosis, we used a flexible parametric model stratified by histotype. The models included period of diagnosis as an ordinal variable, stage group (I–II, III–IV), natural splines of age at diagnosis, and region of residence. All covariates were considered also as time-varying coefficients, with one degree of freedom for period of diagnosis, three for stage and age, and two for region. Excess hazards were predicted from models for females with a median age at diagnosis (63.4 years), diagnosed in the most common area (Central Area) and for each calendar period, each histological subtype and stage group (I–II, III–IV).

Results
A total of 62,450 invasive ovarian cancer cases were included. A total of 74% of cases were of high-grade serous histology and 66% of cases were diagnosed at stage III or IV. 

Excess hazard rates differed substantially by histotype and stage, and also within histotype by time since diagnosis (Figure, Top: stages I/II; Bottom: stages III/IV). Among individuals with the predominant high-grade serous tumors diagnosed at stage III or IV, the excess mortality hazard was notably lower in the most recent diagnosis period (2020-2023), as compared to the earliest period (2010-2012) (Figure, Bottom). A similar, though more modest, pattern suggesting improved survival in more recent years was observed for high-grade serous tumors diagnosed at stages I or II and endometrioid tumors. Survival patterns over time differed among the other histotypes, with more modest improvements across the evaluated time periods.

Conclusions and Implications
Ovarian cancer survival improved over the evaluated periods, in particular for the predominant high-grade serous and endometrioid subtypes. These findings likely reflect a combination of improved surgical care and the introduction of targeted therapies. Continued efforts are needed to develop and refine histotype-specific treatment regimens for individuals diagnosed with ovarian cancer, and toward primary prevention and earlier detection. 

Excess mortality hazard rates substantially by histotype and stage, and within histotype by time since diagnosis (Figure, Top: stages I/II; Bottom: stages III/IV)