Comprehensive assessment of incidence, mortality, and lifetime risk of ovarian cancer: a global study with forecasts to 2050

Ovarian cancer remains one of the most lethal gynecologic malignancies worldwide, primarily due to its insidious nature and frequent late-stage diagnosis [ 1 ]. The 5-year survival rate remains low, and research has shown that the 5-year cause-specific survival for cases diagnosed at stages III and IV, which make up 80% of total incidents, is only 41% and 20%, respectively [ 2 ]. Despite advances in surgery, chemotherapy, and targeted therapies such as poly(ADP-ribose) polymerase inhibitors, the overall prognosis for ovarian cancer has seen only modest improvement over the past few decades [ 3 ]. According to GLOBOCAN 2022, ovarian cancer accounted for over 300,000 new cases and 200,000 deaths globally [ 4 ], posing a substantial challenge to women’s health. Multiple risk factors contribute to the genesis and development of ovarian cancer. Research has demonstrated that genetic factors, such as BRCA1 and BRCA2 gene mutations, significantly increase the lifetime risk of developing ovarian cancer [ 5 6 7 ], supplemented by factors including family history, nulliparity, endometriosis, and hormone replacement therapy. Conversely, factors such as oral contraceptive use, full-term pregnancy and breastfeeding, and tubal ligation are associated with risk reduction [ 8 9 10 ]. Emerging evidence has also highlighted the role of lifestyle, such as diet and physical activity, in modifying ovarian cancer risk [ 11 12 ]. The global burden of ovarian cancer is not evenly distributed. High-income regions and countries such as Northern Europe usually report higher incidence rates, likely due to population aging and better diagnostic capabilities [ 13 14 ]. Studies have reported that the incidence and mortality rates of ovarian cancer are strongly correlated with the human development index, standard of living, educational level, lifestyles, cultural beliefs, and environmental exposures [ 15 ]. Socioeconomic disparities further compound these challenges due to limited healthcare infrastructure, insufficient access to early diagnosis and treatment, and a lack of awareness of symptoms and risk factors for ovarian cancer. Given the substantial burden of ovarian cancer, this study leverages GLOBOCAN 2022 data to evaluate the current incidence, mortality, and distribution patterns of ovarian cancer worldwide. It further seeks to analyze the lifetime risk and future trends of ovarian cancer across regions and socioeconomic contexts, thereby providing critical evidence to guide public health planning and decision-making.

There were over 300,000 new cases and 200,000 deaths of ovarian cancer globally in 2022. Fig. 1 reveals significant regional disparities in ASIR ( Fig. 1A ) and ASMR ( Fig. 1B ). The highest incidence rates were clustered in Eastern and Northern Europe, North America, and parts of Asia (>8.2 per 100,000 women), and the lowest in parts of Africa and Latin America (0.7–4.1). Mortality was most pronounced in Eastern Europe and Central Asia (4.5–6.3). Notably, the data highlighted a critical disparity: sub-Saharan Africa endured high mortality rates against a backdrop of low incidence, while Western Europe and Northern America achieved lower mortality (0.8–3.3) despite higher incidence. ASIR, age-standardized incidence rate; ASMR, age-standardized mortality rate; ASR, age-standardized rate. To further contextualize these distribution differences in ASIR and ASMR, we stratified countries based on geographic region (as defined by the World Bank) and SDI levels ( Fig. 2 ). Europe and Central Asia recorded the highest average incidence rates (8.28), followed by North America (8.27), Southeast and East Asia and Oceania (7.66), South Asia (5.72), North Africa and the Middle East (5.55), Latin America and the Caribbean (5.31), and Sub-Saharan Africa (4.81) ( Fig. 2A ). In addition, mortality was disproportionately different relative to incidence across regions ( Fig. 2B ). Despite a lower incidence rate, regions such as Sub-Saharan Africa and South Asia exhibited a comparatively higher mortality rate. Conversely, North America, and Southeast Asia, East Asia, and Oceania demonstrated a lower average mortality relative to their incidence rate. Specifically, the mortality-to-incidence ratio (MIR) of the Sub-Saharan African region was 0.77, while North America showed a MIR of 0.50, as shown in Table S1 . Notably, variations were seen in incidence and mortality rates across intra-regions ( Fig. 2A and B ). Take Europe and Central Asia as an example, Tajikistan’s incidence rate was under 4, while several other countries reported rates exceeding 12. Moreover, the average mortality rate in this region was 4.85, yet two countries diverged sharply, with one below 2.5 and another above 7.5. A similar disparity was evident in Sub-Saharan Africa, where the average mortality was 3.71, but one country reported mortality rates exceeding 7.5. These findings indicate that the disparities of ovarian cancer burden may not only be determined by incidence but also by inequalities in early detection and treatment. When stratified by SDI ( Fig. 2C and D ), high-middle SDI countries endured the highest average incidence (8.21) and mortality (4.96) rates, followed by high (7.93 and 4.51), middle (5.83 and 3.95), low-middle (5.12 and 3.72), and low (4.62 and 3.59) SDI countries. Of note, 30 of 41 (73%) high SDI countries had ASMR below 5, which was lower than the corresponding ASIR (around 8). In contrast, mortality rates in low SDI countries were relatively higher compared with their incidence rates, reflected by an MIR of 0.78, as shown in Table S2 . These variations likely reflected differences in diagnostic capacity, life expectancy, and data quality. Data source: GLOBOCAN 2022 (version 1.1, released 8 February 2024). ASIR, age-standardized incidence rate; ASMR, age-standardized mortality rate; SDI, sociodemographic index. Fig. 3 ranks the top 20 countries by disparity between ASIR and ASMR for ovarian cancer. For women of all ages ( Fig. 3A ), 19 out of these 20 countries were high or high-middle SDI countries, with Latvia ranking the top (incidence, 15.33; mortality, 7.32), followed by Brunei Darussalam (13.16 vs. 6.92), Lithuania (12.77 vs. 6.78), Japan (8.74 vs. 2.97), and Russian Federation (11.47 vs. 6.14). In contrast, Samoa (14.21 vs. 9.29) presented a notable exception as the only low SDI country (ranked 9th). A smaller number of low SDI countries in this list indicates mortality rates are closer to incidence rates in these countries, underscoring a poorer prognosis that likely reflects more limited healthcare resources in developing countries compared to developed countries. For women under 40 years ( Fig. 3B ), both incidence and mortality rates remained low across countries and SDI levels. Incidence rates in the top 20 countries in this group ranged from 2.12 to 4.98, and mortality rates from less than 0.01 to 0.93. Of these 20 countries, 15 were high or high-middle SDI countries, four were middle SDI, and one was low SDI. Brunei Darussalam, a high-middle SDI country, recorded the widest disparity (4.98 vs. 0.7), followed by the middle SDI nation of Fiji (4.76 vs. 0.68). The low SDI country, Timor-Leste, also appeared on the list, despite a narrower gap. For women aged 40 years and older ( Fig. 3C ), ovarian cancer incidence and mortality were substantially elevated, with Latvia (41.91 vs. 21.98) ranking the top, and Serbia (25.76 vs. 15.60) ranking 20th, indicating that the ovarian cancer burden is predominantly concentrated in the older age group. Notably, 19 out of the top 20 countries were high or high-middle SDI countries, and one was a low SDI country. This list closely mirrored that of the women of all ages group, but showed more pronounced disparities compared to the women under 40 years of age group. These findings suggest that survival disparities are most evident in older women, directly pointing to inequalities in access to advanced diagnostics and quality of medical care. Data source: GLOBOCAN 2022 (version 1.1, released 8 February 2024). ASIR, age-standardized incidence rate; ASMR, age-standardized mortality rate; ASR, age-standardized rate; SDI, sociodemographic index. An asterisk (*) indicates countries classified as low or high SDI. Fig. 4 illustrates the overall trend of LRI and LRM decreased from 2000 to 2021, and these trends varied considerably across the study period, global regions and SDI levels. At the global regional level ( Fig. 4A and B ), the LRI and LRM remained highest throughout the period in North America, and LRI decreased from approximately 1.79% in 2000 to 1.26% in 2021 and LRM from roughly 1.23% to 0.89%. Followed by Europe and Central Asia, the LRI decreased from 1.48% to 1.25%, and the LRM decreased from 1.06% to 0.92%. Latin America and Caribbean (LRI, 0.86% to 0.89%; LRM, 0.65% to 0.63%), and Southeast Asia, East Asia, and Oceania (LRI, 0.66% to 0.69%; LRM, 0.43% to 0.41%) exhibited a moderate risk among the seven global regions, with a relatively stable LRI and LRM over the past 20 years. In contrast, South Asia (LRI, 0.54% to 0.72%; LRM, 0.41% to 0.53%), Sub-Saharan Africa (LRI, 0.47% to 0.65%; LRM, 0.37% to 0.51%), as well as North Africa and Middle East (LRI, 0.44% to 0.56%; LRM, 0.33% to 0.41%) demonstrated a rising LRI and LRM over the same period, highlighting a growing ovarian cancer burden. Similarly, when stratified by SDI ( Fig. 4C and D ), the highest LRI and LRM were observed in high SDI countries, despite a decline in both measures (LRI, 1.6% to 1.2%; LRM, 1.1% to 0.85%). High-middle SDI countries followed a similar declining trend (LRI, 0.99% to 0.88%; LRM, 0.70% to 0.62%). In contrast, middle (LRI, 0.58% to 0.68%; LRM, 0.41% to 0.44%), low-middle (LRI, 0.53% to 0.70%; LRM, 0.38% to 0.50), and low (LRI, 0.46% to 0.63%; LRM, 0.37% to 0.49%) SDI countries experienced gradual increases in both LRI and LRM. Notably, the substantial overlap of 95% confidence intervals (The shaded ribbons in Fig. 4 ), particularly after 2015, suggests that differences across regions and SDI levels were less pronounced and should be interpreted with caution. GBD, Global Burden of Disease; LRI, lifetime risk of incidence; LRM, lifetime risk of mortality; SDI, sociodemographic index. Then we analyzed the AAPC for LRI or LRM from 2000 to 2021 by countries, global regions and SDI levels ( Fig. 5 , Tables S3 , S4 , S5 , S6 ). At the national level, Georgia presents the highest AAPC in both LRI (6.57%) and LRM (7.19%) of ovarian cancer ( Fig. 5A and B ), followed by Turkmenistan (LRI, 5.6%; LRM, 5.6%), Kyrgyzstan (4.1% and 4.1%), the United Arab Emirates (3.3% and 3.9%) and Cambodia (3.1% and 2.9%). Conversely, decreases were noted in Estonia (−3.3%), Australia (−3.2%), Sweden (−2.6%), New Zealand (−2.5%), and Denmark (−2.2%) in AAPC for LRI. It is noteworthy that Denmark ranked among the top 5 countries for decline in LRI, but not for LRM, whereas Greenland showed a decline in LRM (−2.2%), ranking fourth globally. These divergent patterns across countries may reflect differences in healthcare infrastructure, accessibility of screening and diagnostics, as well as variability in the quality of cancer registry data. At the regional level ( Fig. 5C and D ), Sub-Saharan Africa showed the most pronounced upward trends in LRI and LRM, with both AAPCs exceeding 1.5%, followed by South Asia (1.5% and 1.4%), and North Africa and the Middle East (1.2% and 1.1%). In contrast, Latin America and the Caribbean, as well as Southeast Asia, East Asia and Oceania demonstrated moderate increases in LRI but very slight decreases in LRM. Europe and Central Asia remained slightly decreased in LRI (−0.8%) and LRM (−0.7%), whereas North America showed a notable decrease in both LRI (−1.9%) and LRM (−1.8%). When stratifying by SDI level ( Fig. 5E and F ), low and low-middle SDI countries presented obvious increases in both LRI and LRM, with AAPCs exceeding 1%, followed by middle SDI countries (0.8% and 0.5%). In contrast, high-middle (−0.6% and −0.7%) and high (−1.5% and −1.4%) SDI countries showed notable decreases, suggesting a growing global disparity in the ovarian cancer burden between low and high SDI countries. To assess the robustness of the estimated AAPCs, we performed sensitivity analyses. In certain countries or regions, excluding the earliest or latest data points led to modest variations in the AAPCs ( Figs. S1 , S2 , S3 , S4 , S5 , S6 ); however, the fundamental trends remained consistent. Additionally, the rolling-window analysis further demonstrated similar periods of acceleration and deceleration regardless of methodological approach ( Tables S7 and S8 ). Fig. 5 also visualizes this consistency, with the error bars depicting the range of 5-year AAPC estimates. Data source: Global Burden of Disease Study 2021. AAPC, average annual per cent change; APC, annual percent change; ASIR, age-standardized incidence rate; ASMR, age-standardized mortality rate; SDI, sociodemographic index. Based on projected demographic shifts, the global burden of ovarian cancer is expected to increase substantially from 2022 to 2050, with 477,000 new cases (a 46.9% increase) and 337,000 deaths (a 62.7% increase) projected for 2050 ( Fig. 6 ). Of note, both incidence and mortality are expected to rise across all SDI levels, but the magnitude varies dramatically. The largest increases are anticipated in low SDI countries (incidence: +153.8%; mortality: +165.2%), followed by moderate increases in low-middle (+88.6% and +104.9%) and middle (+60.0% and +81.2%) SDI countries. In contrast, high-middle (+14.3% and +35.0%) and high (+21.8% and +36.4%) SDI countries are projected to see more modest changes ( Fig. 6A and B ). When stratified by age, for women under 40 years old, increases are expected in low (+103% and +102.3%), low-middle (+20.8% and +21.1%), and middle (+2.1% and 3.7%) SDI countries, and declines are projected in high-middle (−24.2% and −21.9%) and high (−10.9% and −9.9%) SDI countries ( Fig. 6C and D ). For women aged 40 years and older, the projected increases are pronounced in low (+174.4% and +177.0%) and low-middle (+102.9% and +112.3%) SDI countries. In contrast, modest increases are anticipated in high-middle (+18.5% and +36.7%) and high (+23.9% and +36.9%) SDI countries ( Fig. 6E and F ). Data source: Global Burden of Disease Study 2021, Institute for Health Metrics and Evaluation. SDI, sociodemographic index.

Ovarian cancer is one of the most frequently diagnosed gynecologic malignancies, with incidence and mortality rates varying greatly across countries and regions. This study describes the latest epidemiology of ovarian cancer, including incidence, mortality, and lifetime risk, and projects its burden to 2050 at the global, regional, and national levels. We observed that developed countries bear a higher incidence burden, yet developing countries face a disproportionately higher mortality burden, indicating critical disparities in healthcare access. While the overall LRI and LRM from ovarian cancer have declined over the past two decades, this trend varies by regions and SDI levels. Alarmingly, we project a substantial increase in the future global burden of ovarian cancer, underscoring the urgent need for global action. According to the GBD data, the global incidence and mortality associated with ovarian cancer increased from 1990 to 2019, reaching 294,422 and 198,412 cases, respectively. These numbers rose to over 300,000 new cases and 200,000 deaths in 2022 [ 4 20 ], and are projected to grow substantially to 477,000 and 337,000 by 2050. This trend indicates a huge increase in the global burden of ovarian cancer in the future. Analysis of age-standardized rate (ASR) reveals a similar pattern. The ASIR increased from 152.1 per 100,000 in 1990 to 286.1 in 2017, and the ASMR rose from 95.5 to 176.0. However, the distribution of these rates is highly uneven across regions: countries in Western Europe experience the greatest decreases, while nations in Latin America and the Caribbean have seen the most pronounced increases [ 21 ]. These findings were consistent with the results of our analysis. Besides, our analysis showed these disparities were strongly associated with SDI levels, with the highest ASR observed in high SDI countries. This correlation can be explained by variations in healthcare infrastructure. Extensive research has established that access to early screening, advanced cancer treatments, nutritional support, and rehabilitation services is essential for improving the survival of ovarian cancer patients. Developed countries tend to provide better comprehensive medical care, which may explain their relatively higher incidence rate but lower mortality rate. In contrast, the limited health care access in developing countries often results in later diagnosis and higher MIRs [ 14 22 23 ]. Global trends in ovarian cancer have shown a rise in incidence and mortality over the past 2 decades, while the overall trends of LRI and LRM were decreasing, and they varied hugely across regions and SDI levels. For instance, high SDI countries saw a marked decline from 2000 to 2021, whereas low SDI countries experienced a slight increase in LRI and LRM. This pattern aligns with the observed AAPCs in ASIR and ASMR, which decreased in developed countries like those in Western and Northern Europe but increased in many developing regions, including South Asia, North Africa, and the Middle East. These disparities are driven by multiple factors. Beyond unequal access to cancer care, genetic susceptibility may play a role. For example, the prevalence of BRCA1/2 mutations is higher in women of African descent (15.6%) compared to Western populations (12.2%) [ 24 ], which may contribute to the rising trend of AAPCs in some African regions. Moreover, the stark difference in outcomes is further supported by survival data: the 5-year survival rate in developed countries was 36%–46% (2010–2014) [ 25 26 ], while this rate in developing countries can be as low as 18.6% (India) to the mid-30s (Brazil, Thailand), with an overall estimated range of 20%–35% [ 26 ]. Notably, improvements in survival rate for ovarian cancer in developed countries ranged from 4.4% to 9.8% from 1995–1999 to 2010–2014, and were much lower compared to that of rectal cancer (9.5% to 21.0%) during the same period [ 25 ]. The high mortality rates and slow improvements for ovarian cancer were largely attributed to the lack of effective early screening methods, leading to most cases being diagnosed at a late stage. Our study also demonstrated that LRI and LRM for ovarian cancer rose more rapidly after age 40, highlighting the need for the development of effective screening methods. Projections based on 2022 data indicate a sharp rise in the global burden of ovarian cancer by 2050, with a 46.9% increase in incidence and a more pronounced 62.7% increase in mortality. Critically, this burden is strongly correlated with SDI levels. Low SDI countries are projected to bear a disproportionately higher share of new cases and deaths, a trend consistent with other malignancies [ 27 28 ], highlighting widening disparities in the future. Consequently, it is necessary to call for global attention to reduce the disparities of ovarian cancer burden and facilitate access to quality cancer care and treatment. There were several limitations of this study. First, the accuracy of lifetime risk estimates is contingent upon the quality and completeness of the underlying cancer registry data. However, the reliability of these registries varies substantially across regions, with data from areas such as Africa and South America being particularly limited. Second, the observed associations between ovarian cancer metrics (incidence, mortality, lifetime risk) and SDI levels may reflect international disparities in diagnostic capacity and access to medical care. Therefore, underdiagnosis of ovarian cancer in lower-resource settings may be an important factor that compromises the precision of lifetime risk estimations. Third, GLOBOCAN and GBD databases provide age information, but lack data on histology and clinical stage. This precludes an aggregated analysis by these factors and limits the clinical interpretability of our findings.

We obtained ovarian cancer incidence and mortality data from 2 major global databases: GLOBOCAN 2022 and the Global Burden of Disease (GBD) 2021 study. From GLOBOCAN 2022 (International Agency for Research on Cancer, IARC) [ 16 ], we downloaded country-level age-standardized incidence rates (ASIR) and age-standardized mortality rates (ASMR) in 2022, as well as the projected new cases and deaths for ovarian cancer among women of all ages. From the GBD 2021 database [ 17 ], accessed via the Global Health Data Exchange, we extracted age-band data on ovarian cancer incidence and deaths for females across standard 5-year age groups (5–9, 10–14, 15–19, 20–24, …, 80–84, and 85+ years), along with corresponding all-cause mortality rates for the same age ranges. The study encompassed data from 185 countries and territories, supplemented with country-level metadata, including World Bank income classifications and regional groupings ( https://www.worldbank.org/ ), to facilitate comparative analysis. The datasets were used to generate both descriptive and analytical insights. First, we visualized ovarian cancer incidence and mortality patterns across countries, stratified by World Bank region, sociodemographic index (SDI) level, and age group (<40 and ≥40 years). Second, the data were used for further analyses, including lifetime risk estimation, temporal trend assessment, and future burden projection. In this study, we estimated two key metrics: the lifetime risk of incidence (LRI) for ovarian cancer, defined as the probability of being diagnosed with the disease during a person’s lifetime (excluding multiple primary cancers), and the lifetime risk of mortality (LRM) for ovarian cancer, defined as the probability of dying from the disease over a lifetime. Traditional methods for estimating these risks included the cumulative rate, cumulative risk, and the life table approach. We adopted the Age-Mortality-Population (AMP) method [ 18 19 ], a life table-based approach that further adjusts for the occurrence of multiple primary cancers in the population, offering more accurate and realistic estimates. Mathematically, for each 5-year age group i , the AMP-lifetime risk algorithm takes as inputs: R i , the age-specific cancer incidence rate (number of new cancer cases per 100,000 persons per year); M i , the all-cause mortality rate (deaths from any cause per 100,000 persons per year); D i , the cancer-specific mortality rate (deaths from the same cancer per 100,000 persons per year); and w i , the width of the age interval (usually 5 years). At the start of each interval, the probability of being alive and cancer-free, denoted S i , is computed recursively as the product of the survival probabilities from all preceding intervals. The total exit rate from the cancer-free state is given by λ i = ( R i + M i − D i )/100,000, representing the combined hazard of developing cancer or dying from other causes. The probability of developing cancer within interval i is then approximated as The overall lifetime risk of developing (or dying from) cancer is obtained by summing the interval-specific probabilities across all age groups, This formulation accounts for competing mortality, ensuring that individuals who die of other causes are removed from the population at risk in subsequent age intervals. When R i is replaced by D i , the same framework yields the lifetime risk of mortality. The AMP method, therefore, provides internally consistent and life table–based estimates of the probability that an individual will develop or die from ovarian cancer during her lifetime, under the assumption that the age-specific incidence and mortality rates of the reference year remain constant over time. To quantify uncertainty, we recomputed the AMP estimates using the lower and upper bounds of the GBD-reported 95% uncertainty intervals for all three rate inputs (incidence, ovarian cancer-related mortality, and all-cause mortality). These runs yielded the lower and upper uncertainty limits of the lifetime risk estimates presented in this study. The resulting 95% uncertainty intervals reflect the propagation of uncertainty from the underlying GBD estimates. Temporal trends in age-standardized risks of ovarian cancer incidence and mortality were examined using log-linear regression models. For each country, we fitted a single-slope model of the form: Where y t represents the lifetime risk in year t . The regression coefficient ( β 1 ) quantifies the constant annual rate of change on the logarithmic scale and is converted to the average annual per cent change (AAPC) as: To further assess the stability of the estimated AAPCs, several sensitivity analyses were performed, including truncation of the first or last 2 years of observation, alternative model specifications (linear versus log-linear scales), and smoothing using a 3-year moving average. These analyses yielded consistent results, supporting the robustness of the observed trends. Given the absence of distinct join points, we additionally explored short–term fluctuations using a 5-year rolling–window approach. For each consecutive 5-year period, the mean annual per cent change was calculated as: and the maximum and minimum values across all windows were extracted to identify periods of most rapid increase or decline. We obtained estimates and projections for ovarian cancer incidence and mortality from the GLOBOCAN database for the period 2022–2050. GLOBOCAN projections were generated by applying the age-specific incidence and mortality rates observed in 2022 to population projections from the United Nations World Population Prospects, under the assumption that age-specific rates remain constant over time. Thus, these projections reflect expected demographic changes, principally population growth and ageing, rather than changes in the underlying cancer risk or survival. We visualized the temporal trends in the estimated and projected numbers of new cases and deaths, and calculated the percentage change between years using the formula %Δ = ( N t 2 − N t 1 )/ N t 1 × 100%, where N t1 and N t2 denote the numbers of cases or deaths at the beginning and end of the comparison period, respectively. We stratified analyses by SDI and age group (<40 vs. ≥40 years) to capture variations in the projected burden of ovarian cancer across different socioeconomic and demographic profiles. All data processing, lifetime risk estimation, and visualizations were performed using R (version 4.4.0; R Foundation for Statistical Computing, Vienna, Austria). The analyses employed key packages including dplyr , ggplot2 , patchwork , and readr . To facilitate geographic or socioeconomic comparisons, country names were harmonized across datasets and mapped to relevant group classifications. This study employed previously published, de-identified human data from the GBD study, where ethics approval and consent to participate were not required for secondary data analysis.