Estimating global cancer survival and mortality from 1990 to 2050: a simulation-based analysis of 17 cancers.

Abstract

Cancer survival has improved in many countries in recent decades, but large disparities remain globally. Understanding the effects of workforce shortages, stage at diagnosis, and broader health system barriers on poor survival is important for the design and evaluation of policy interventions to improve cancer survival in diverse settings. We aimed to develop a model to estimate trends in cancer survival (overall and by stage), accounting for these factors. We developed the Global Cancer Workforce microsimulation model to simulate 17 cancer types (oral, nasopharynx, oesophagus, stomach, colon, rectum, anus, liver, pancreas, lung, skin melanoma, breast, cervix uteri, ovary, prostate, bladder, and brain and CNS) in 200 countries and territories from 1990 to 2050, accounting for trends in stage-specific survival and health system factors, including 14 specific workforce personnel types relevant to cancer survival. We calibrated the survival module to empirical data on population-based 5-year net survival, and estimated survival in each country by cancer and overall (pooled and standardised for country-specific incidence of different cancer types). We also estimated stage-specific cancer survival, and cancer mortality rates based on both observed (diagnosed) and total (undiagnosed and diagnosed) cancer cases. Between 2025 and 2050, global pooled 5-year net survival was projected to remain stable (47·6% [95% uncertainty interval (UI) 45·9-49·1] to 47·7% [46·0-49·4]), with persistent disparities by income group and geographical area. In 2050, pooled survival was projected to be lowest in Africa (34·4% [32·6-36·3]) and Asia (38·7% [36·0-40·6]), and highest in North America (63·9% [58·1-66·6]) and Oceania (70·4% [67·2-73·7]). Standardising survival for country-specific incidence of different cancer types resulted in decreased estimates for all world regions except eastern Asia, which has a higher proportion of cancer types with poor survival. Large disparities in survival by geographical region also existed within each cancer type. For example, in 2025, estimated 5-year net survival for melanoma ranged from 39·0% (22·2-57·2) in Africa to 91·7% (85·1-95·7) in North America. Global cancer mortality in 2025 was estimated to be 90·0 per 100 000 population (95% UI 79·6-97·2) based on diagnosed cases, which increased to 188·9 per 100 000 population (183·4-195·5) when total cancer cases were considered, with substantially higher estimates compared with those based on observed cases in settings such as Africa and south Asia. Disparities in global cancer survival are large and are expected to persist on current trends without transformative policy action. Our modelling framework provides comprehensive, country-level projections of survival and mortality for multiple cancers, accounting for demographic and epidemiological trends, as well as health system barriers and workforce constraints. American Cancer Society and Breast Cancer Research Foundation.