National and regional projections of gynaecological cancers in Aotearoa New Zealand: a projection model to 2045 highlighting an opportunity for collective action

Gynaecological cancers are increasing globally, with incidence rates rising across multiple cancer types. In 2022, an estimated 1.47 million new cases and 680,000 deaths from gynaecological cancers were recorded worldwide, and both figures are projected to grow substantially over the coming decades.^1,2 The incidence of uterine cancer has risen sharply in high-income countries, driven in part by population ageing, changes in reproductive patterns and increasing prevalence of various risk factors including diabetes and excess body weight.³ Ovarian cancer incidence has also grown globally, with the number of new cases nearly doubling between 1990 and 2021.²

In Aotearoa New Zealand, similar trends are emerging. Recent data show a marked increase in uterine cancer incidence, which has risen by nearly 60% over the past decade, and this rate of increase is among the highest reported for any country.^4–7 Higher cancer incidence and lower survival for Pacific peoples is increasingly being recognised as an area that needs focussed attention.⁸ These patterns signal a need for more data to support targeted planning and service delivery at both national and regional levels, particularly as the patterns of gynaecological cancer become more unevenly distributed across age and ethnic groups.

As the only gynaecological cancer with a national screening programme, the incidence rates of cervical cancer are projected to decline further with the continuing impact of human papillomavirus (HPV) vaccination and HPV-based screening.⁹ The World Health Organization (WHO) has also set elimination targets for cervical cancer.¹⁰ However, for other cancers such as uterine cancer, incidence is rising, there are few population-level strategies in place for early detection or prevention and there is no clear screening intervention at present.⁵ The patterns of incidence are likely to affect some regions in Aotearoa New Zealand more than others given the variations in demographic makeup and may widen existing inequities if not planned for, particularly given the capacity constraints on both general gynaecology and gynaecological oncology services nationally (including the additional colposcopy demands related to HPV primary screening).

Existing cancer projections in Aotearoa New Zealand are generally presented only at the national level and do not include cancer-specific breakdowns by ethnicity.¹¹ While nationally aggregated figures are useful for high-level policy decisions, they are often too general to guide service delivery and planning at the regional or district level. This limits the ability to understand how future demographic shifts, such as the projected increase among Pacific peoples, will shape cancer incidence, and it constrains hospital services, cancer planners and regional health planners in anticipating and responding to emerging pressures.

This paper partly addresses this gap by presenting both national and regional projections of gynaecological cancer incidence to 2045. Using historical cancer registry data and population projections, we estimate projected case counts for uterine, cervical, ovarian and other gynaecological cancers. Projections are provided at the national level and disaggregated by region and ethnicity to support targeted planning and identify where a more tailored or equity-focussed response may be needed.

Methods

Data sources

National cancer registration data for Aotearoa New Zealand from 2001 to 2022 were used to project gynaecological cancer incidence. The analysis covered cervical (International Classification of Diseases 10 [ICD-10] C53), uterine (C54 and C55), ovarian and fallopian tube (C56 and C57.0) and other gynaecological cancers including vulva (C51), vagina (C52), placenta (C58) and unspecified female genital organs (C57 excluding C57.0). Data were aggregated by calendar year, 5-year age groups, prioritised ethnicity (Māori, Pacific, Asian, Other) in line with Health Information Standards Organisation 10001:2017 protocols¹² and the four Health New Zealand – Te Whatu Ora regions. Population estimates for 2001 to 2022 and official population projections for 2023 to 2045 with the same stratification were sourced from Statistics New Zealand. These projections were used as denominators in the projection period so that expected future population growth, ageing and ethnic composition are reflected in the projected case numbers.

Modelling approach

An age-period-cohort (APC) model was developed to estimate incidence trends. Separate Poisson generalised linear models with a log link were fitted for each cancer site, with case counts as the outcome and log population as an offset. Predictors included age group, calendar year, birth cohort, prioritised ethnicity (see limitations section for discussion on ethnicity) and region.

The APC structure was broadly based on the Norwegian Prediction Model (Nordpred),¹³ which captures period and cohort effects alongside age. Despite the similarities, a new model was built to better reflect Aotearoa New Zealand data and policy needs, allowing stratification by ethnicity, region and age group, while retaining sparse or zero case counts. No interaction terms were included to reduce overfitting and maintain interpretability. Bootstrapping (described below) helped address instability in small strata.

To emphasise recent data, cancer registrations were weighted by year, with weights raised to the power of 1.5. Incidence rates were forecast to 2045 by extending the period and cohort components beyond 2022, and projected case numbers were obtained by applying these forecast rates to the corresponding Statistics New Zealand projected populations within each stratum. Analyses were conducted in R (version 4.4.3; R Core Team, 2024).

Bootstrap procedure

Uncertainty was modelled using non-parametric bootstrapping. For each cancer site, 1,000 bootstrap samples were generated by resampling historical incidence data and adding random variation to the population denominator (mean=1, standard deviation [SD]=0.01). A new Poisson model was fitted for each sample and applied to future population projections (2023–2045). For each stratum, median projected counts and 95% uncertainty intervals (UI) (2.5th–97.5th percentiles) were calculated across iterations.

This approach produced distributions of plausible future trends, reflecting variability in historical data and providing a more realistic measure of uncertainty than a single fitted model. To account for increasing uncertainty in population projections, additional perturbation was applied stepwise (SD=0.005 for 2023–2025, 0.01 for 2026–2030, 0.015 for 2031–2035, 0.02 for 2036–2040 and 0.025 for 2041–2045). Projected case counts were obtained by exponentiating the linear predictor from each fitted model.

Age-standardised rates (ASRs)

ASRs were calculated for both observed and projected data from 2001 to 2045. Rates were produced separately for each cancer site and for total gynaecological cancer, stratified by year, Health New Zealand region and prioritised ethnicity. The WHO 2001 population standard was used as the standard population.

Classification changes

Since around 2014, increasing adoption of the fallopian tube origin hypothesis for high-grade serous carcinomas has influenced coding practices.^14–16 Many cancers previously coded as ovarian (C56) are now classified as fallopian tube (C57.0), creating an apparent decline in ovarian incidence and a rise in C57.0 that largely reflects coding shifts rather than real changes. To account for this in our model, ovarian cancer was defined here as C56 plus C57.0, consistent with current understanding of disease origin and to maintain comparability over time. Nonetheless, some misclassification may persist, and projections assume coding consistency, so caution is warranted when interpreting site-specific trends, particularly ovarian cancer.

An increase was also observed from 2014 in unspecified or overlapping gynaecological sites (C57.7–C57.9). Unlike C57.0, which was combined with ovarian cancers, these were grouped as “other gynaecological cancers” and retained as coded. Their rise may reflect changes in diagnostic or coding practices when the primary site was uncertain or involved multiple locations and a primary cancer site could not be established.

Study approval

This study was considered low risk and was out of scope for Health and Disability Ethics Committee review. Locality authorisation was granted by Health New Zealand – Waitematā District, Research & Knowledge Centre (Approval Code: WAI20367).

Results

National projections

During 2020–2022, an average of 1,375 gynaecological cancers were diagnosed annually (Table 1): uterine cancers comprised just over half (717), followed by ovarian/fallopian tube (327), cervical (169) and other gynaecological cancers (162). The combined ASR was 36.6 per 100,000 (95% confidence interval [CI] 35.5–37.8), with uterine cancer having the highest site-specific rate (18.6; 95% CI 17.8–19.4).

By 2035, annual cases are projected to rise 44% to 1,982 (95% UI 1,854–2,120). Growth is driven by uterine (+59%) and other gynaecological cancers (+90%), while cervical (+6.5%) and ovarian/fallopian cancers (+5.5%) increase only slightly. The overall ASR rises to 40.0 (95% UI 36.2–44.1), with increases for uterine (23.0) and other cancers (5.1), but declines for cervical (5.1) and ovarian/fallopian cancers (6.7).

By 2045, nearly 2,500 cases are projected to occur annually (95% UI 2,263–2,774), an 82% increase from baseline. Uterine cancers more than double to 1,506 (95% UI 1,310–1,739), and other gynaecological cancers rise 184% to 460 (95% UI 358–612). Cervical and ovarian/fallopian cancers show only modest increases in numbers (~5%) but declining ASRs. The overall ASR reaches 44.2 (95% UI 38.9–50.3), with uterine cancer increasing 46% to 27.1 and other cancers 78% to 6.4, while cervical and ovarian/fallopian ASRs fall by 19% and 33%, respectively.

View Table 1–3, Figure 1.

Ethnic-specific projections

For Māori women, annual gynaecological cancer cases (Table 2) are projected to more than double from 214 in 2020–2022 to 497 by 2045 (95% UI 449–553), mainly from uterine cancers (118 to 308; ASR 27.3 to 39.6) and other gynaecological cancers (21 to 82; ASR 5.0 to 9.4). ASRs for cervical and ovarian/fallopian cancers show slight declines.

Among Pacific women, total gynaecological cancer cases are projected to rise from 165 to 391 (95% UI 340–449), with the overall ASR increasing from 97.3 to 122.3, the highest of any group. Uterine cancers more than double (126 to 312; ASR 74.3 to 97.5). Cervical and ovarian/fallopian ASRs decline modestly, despite small increases in case numbers.

For Asian women, case numbers are projected to more than triple (131 to 423; 95% UI 375–479), but the ASR remains relatively stable (28.4 to 30.0), indicating growth largely due to population growth. Cervical and ovarian/fallopian ASRs decline, while uterine cancers increase slightly (ASR 14.1 to 18.4).

Among European and Other women, cases rise from 866 to 1,185 (95% UI 1,069–1,312), a 37% increase. Uterine cancers grow from 409 to 626 (ASR 13.5 to 19.6), and other cancers increase from 131 to 293 (ASR 3.8 to 6.9). Both cervical and ovarian/fallopian cancers show declining case numbers and ASRs.

Regional projections

In the Northern Region, cases are projected to rise from 527 in 2020–2022 to 1,033 by 2045 (95% UI 922–1,151), driven largely by uterine (296 to 662; ASR 22.4 to 31.4) and other gynaecological cancers (50 to 168; ASR 3.3 to 6.3). Cervical case numbers remain stable, and ovarian cases increase slightly, though ASRs for both decline.

In Te Manawa Taki, total cases are projected to increase from 267 to 496 (95% UI 447–550). Uterine cancers rise from 134 to 291 (ASR 16.0 to 25.4) and other cancers from 35 to 97 (ASR 3.8 to 6.6). Cervical cancers increase modestly (27 to 36), but ASRs remain largely unchanged.

In the Central Region, cases increase from 256 to 427 (95% UI 385–473), driven by uterine cancers (132 to 253; ASR 17.3 to 25.6) and other cancers (27 to 77; ASR 2.9 to 6.0). Cervical and ovarian rates decline, with case counts staying relatively stable.

In Te Waipounamu, total cases grow from 325 to 544 (95% UI 490–610). Uterine cancers increase from 155 to 302 (ASR 15.9 to 22.8), and other cancers from 51 to 118 (ASR 4.6 to 6.9). Cervical and ovarian ASRs decline, while case numbers remain steady.

Discussion

This analysis projects a substantial rise in gynaecological cancer cases in Aotearoa New Zealand by 2045, with uterine cancers showing the largest absolute growth. These trends are consistent across regions and ethnic groups and align with national and international studies documenting an expected 3% annual increase in endometrial cancer, particularly among younger women and Māori and Pacific populations.^4,17 Excess body weight and diabetes are likely major drivers of changing risk patterns. Excess body weight, strongly linked to endometrial cancer through hormonal and inflammatory pathways, remains highly prevalent, especially among Māori and Pacific women, and is estimated to account for ~40% of cases in Māori and over half in Pacific women.^18–20 Diabetes prevalence is also increasing (~7% annually) and is three times higher in Māori and Pacific peoples, conferring a two- to three-fold increased risk independent of body weight.^21–23

Cervical cancer rates are projected to decline, reflecting the impact of screening and HPV vaccination,⁹ though inequities persist for Māori and Pacific women.²⁴ However, the shift to HPV testing may produce a temporary short-term incidence spike as existing precancers and early-stage cancers are detected.^25,26

While rising incidence partly reflects shifting risk factor profiles, demographic growth is also a major contributor. For example, Asian women are projected to see cases more than triple by 2045, yet ASRs increase only slightly (28.4 to 30.0 per 100,000), suggesting much of the growth in this group is as a result of this demographic change. By contrast, ASRs rise more than 20% in other groups, suggesting stronger risk factor contributions.

Interpretation of ovarian cancer projections requires some caution, given the shift in reclassification of high-grade serous carcinomas to fallopian tube origin,¹⁴ which may have led to lower ovarian rates. Increases in “other” gynaecological cancers (C57.7–C57.9) may also reflect coding and diagnostic uncertainty rather than true increases.²⁷

Regional projections show the largest absolute growth in the Northern Region, reflecting both population growth and higher baseline incidence in Pacific women concentrated in Auckland, while other regions show smaller absolute but still substantial relative increases.

Our findings are broadly consistent with those of Teng et al.,¹¹ who also projected rising uterine cancer and stable ovarian cancer incidence to 2044. Differences include slightly higher growth in uterine cancers and lower cervical cancer incidence in our model, but overall alignment supports the validity of these projections and reinforces the urgent need to address the growing burden of uterine cancer and associated inequities.

A pathway to action—endometrial cancer

The rise of endometrial cancer is increasingly being identified as a global concern.¹ In Aotearoa New Zealand, the trajectory is particularly steep among younger Pacific women.⁶ While these trends are not inevitable, reversing them will require coordinated action across the full prevention and care continuum. Figure 1 illustrates this spectrum, from community awareness to prevention activities targeting key risk factors such as excess body weight and diabetes, to ensuring timely and culturally safe investigation and management of abnormal uterine bleeding and improving access to surgical and non-surgical interventions for endometrial hyperplasia and cancer. Supportive survivorship care is also essential, including fertility preservation.²⁸

Emerging research, both internationally and in Aotearoa New Zealand, is contributing to these efforts by exploring the role of endometrial microbiota in cancer development and investigating biomarkers that may support less invasive diagnostic approaches.^29–32 When addressing the risk associated with excess body weight, it is important to do so in a way that is holistic and pro-equity, as well as anti-stigmatising, and that acknowledges the broader and complex drivers of excess body weight.^18,28

The inequities identified in this study highlight the need for targeted, research-informed strategies, particularly for Pacific populations and in areas with high Pacific populations. Prognosis generally remains good, with 5-year relative survival for localised disease at 90–96%.³³ However, patterns are changing, with earlier age associated with later stage disease and a rise in more aggressive cancers, both of which are associated with poorer survival. Treatment almost always entails a loss of fertility, underscoring the need for early detection and supportive survivorship care.

Cervical cancer elimination

The WHO defines elimination of cervical cancer as an annual rate below four cases per 100,000 women, with 2030 targets of 90% HPV vaccination by age 15, 70% of women screened at ages 35 and 45 and 90% of women with cervical disease receiving treatment.¹⁰ While countries such as Australia have adopted national elimination goals and are projected to reach the threshold within the next decade,³⁴ Aotearoa New Zealand has no explicit goal in place.

Despite not having an elimination goal, our projections indicate that Aotearoa New Zealand is unlikely to meet the elimination threshold even by 2045 based on current trajectories. The projected national ASR in 2045 is 4.7 per 100,000, above the WHO benchmark. By ethnicity, projected 2045 ASRs are 7.3 for Māori women, 6.3 for Pacific women, 4.0 for European/Other women and 3.5 for Asian women. Thus, only Asian women, and possibly European/Other, are on track to reach the elimination threshold, while rates for Māori, Pacific and the overall population remain above the WHO threshold.

Implications for service planning

From a service-planning perspective, the rising number of gynaecological cancers will increase demand across specialist assessment, diagnosis, treatment and follow-up care. The regions should plan capacity increases over the medium term, with clear referral pathways from primary care and timely access to specialist services. Workforce planning will be needed across clinical and diagnostic services, with actions to recruit, retain and upskill staff and to enable care closer to home. Equity must be built in from the outset, including culturally anchored navigation and outreach for Māori and Pacific peoples and active monitoring of access and waiting times. Regional capacity modelling can align with these projections, supported by consistent, high-quality data to forecast demand and capacity across the pathway, monitor access and waiting times, identify bottlenecks and track equity of care, with attention to Māori and Pacific peoples given the uneven distribution of gynaecological cancer incidence.

Strengths

This study provides the first sub-national projections of gynaecological cancer incidence in Aotearoa New Zealand, disaggregated by cancer site, ethnicity and region, through to 2045. Reporting both case counts and ASRs with UIs allows assessment of absolute and relative changes while reflecting statistical variability and a range of future scenarios. Presenting results across Māori, Pacific, and regional populations highlights where the greatest increases are projected to occur, supporting equity-focussed planning and targeted health system responses. Importantly, the model distinguishes between cancer types rather than treating gynaecological cancers as a single category, enabling clearer insights into site-specific trends with different risk factors, screening approaches and treatment pathways. The projections therefore provide timely, policy-relevant evidence to inform prevention, early detection and service delivery strategies in anticipation of rising cancer burden and widening inequities.

Limitations

The accuracy of these projections depends on the validity of the APC assumptions and the stability  of historical trends, yet changes in risk factors, screening uptake, diagnostic practices or coding may alter future incidence in ways not captured. For example, the introduction of primary HPV testing and HPV self-test access is likely to cause a short-term spike in cervical cancer incidence not fully reflected here. The projections should be treated as descriptive tools for planning rather than formal predictive models.

The analysis also relies on registry data coded with ICD-10, which may not fully reflect evolving tumour classification. Diagnostic shifts, such as reclassifying high-grade serous carcinomas from ovarian to fallopian tube origin or increases in overlapping/unspecified sites (C57.7–C57.9), may produce artefactual trends.

Undercounting of some groups is a further limitation, particularly where people identify with multiple ethnicities.^35,36 Use of prioritised ethnicity supports national reporting but may understate the burden for some populations, especially Pacific peoples.³⁷ The broad Level 1 groupings used in this analysis may mask important heterogeneity within the Asian and Pacific populations. More detailed disaggregation—for example, by Chinese, Indian and Other Asian, or by specific Pacific sub-groups such as Samoan, Tongan and Cook Islands Māori—would provide a clearer understanding of differences in cancer risk and projected burden. However, the data required for such analyses, including consistent historical and projected population estimates at this level of detail, are not yet available.

Projections based on APC models are inherently uncertain and should be interpreted as indicative rather than predictive. The models capture long-term incidence patterns and indirectly reflect changes in underlying risk, but they assume that the APC relationships continue in a broadly similar way. The uncertainty intervals presented alongside our projections incorporate both statistical variation in model estimates and uncertainty in future population projections, yet they cannot fully account for all sources of potential error. For context, the projected total number of gynaecological cancers ranges from approximately 1,850 to 2,120 cases in 2035 and from 2,260 to 2,770 in 2045, corresponding to increases of about 35–55% and 65–100% relative to 2020–2022. These ranges illustrate that while the direction of change is robust, there remains some uncertainty around the true magnitude of the projected increase. The projections should therefore be regarded as plausible scenarios that describe the likely direction and relative magnitude of change, providing a useful guide for planning under a range of possible futures.

Conclusion

This study presents the first sub-national projections of gynaecological cancer incidence in Aotearoa New Zealand to 2045 by cancer site, ethnicity and region. Cases are projected to rise substantially, led by rising incidence of uterine cancer, with the largest increases among Māori and Pacific women. These findings highlight the need for equity-focussed prevention, timely diagnosis and culturally responsive care, alongside investment in service capacity to meet future demand.