Racial and socioeconomic disparities in survival improvement of eight cancers

Cancer is a leading cause of death globally. While advancements in cancer treatment have led to significant survival improvements in recent decades, these improvements are not uniformly distributed across all patient populations. Disparities in cancer survival exist based on race, age, sex, and socioeconomic status. Previous research has extensively documented disparities in cancer survival, but disparities in *survival improvement* across these demographic factors remain understudied. Understanding these disparities is crucial for developing targeted interventions and resource allocation to reduce health inequities. This study uses data from the Surveillance, Epidemiology, and End Results (SEER) database to examine survival improvement disparities in eight common cancer types between 2004 and 2018, focusing on race, age, socioeconomic status, and sex.

Existing literature highlights racial disparities in cancer outcomes, showing that Black and Hispanic patients often experience lower cancer-specific survival compared to White patients. Studies have also demonstrated survival differences based on age, sex, and socioeconomic status. While disparities in overall survival have been well-documented, fewer studies have focused on disparities in the *rate* of survival improvement over time. A previous study examining data from 1990 to 2009 found disparities in survival improvements, with variations across age and race (except for ovarian cancer). Given recent advancements in cancer treatment, particularly immunotherapies, a reassessment of these disparities in the context of more recent data is warranted.

This study analyzed data from the SEER 18 registries Incidence-Based Mortality data for cancer patients diagnosed between 2004 and 2018. Eight common cancer types were included. Data quality control measures were implemented, including criteria for reporting source, primary cancer diagnosis, and age at diagnosis (40-85 years). Cancer-specific survival (CSS) was determined using relevant SEER variables. Race, age (categorized into three groups), sex, and annual household income (categorized into three groups) were also recorded. Survival analysis, including multivariable Cox regression, was used to estimate survival disparities within subgroups based on the demographic factors. Patients were grouped into three five-year periods (2004-2008, 2009-2013, and 2014-2018) to analyze changes in survival over time. White race, younger age (40-55), low income (<$60,000), and female sex served as reference groups. Hazard ratios (HRs) and 95% confidence intervals (CIs) were calculated to assess survival improvement disparities across subgroups.

Analysis revealed that White patients had the highest CSS in breast, prostate, and rectal cancers, while Asian patients had the highest CSS in ovarian, liver, lung, and pancreatic cancers. Black and Hispanic patients consistently showed the lowest CSS in several cancers. Older age groups (56-70 and 71-85) showed worse CSS across all cancer sites compared to the youngest group (40-55). High-income patients had higher CSS in some cancers (ovarian, liver, lung, pancreatic) but lower CSS in others (breast, prostate). Female patients generally had higher CSS than male patients. Black patients had significantly lower CSS than White patients in all cancers except lung cancer, with the largest disparity observed in breast cancer. Stratified analysis revealed that while all racial groups experienced improvements in CSS from 2009-2013 and 2014-2018 compared to 2004-2008, black patients with breast, ovarian, and prostate cancer from 2014-2018 still had lower CSS than white patients from 2004-2008. Significant improvements in survival were observed across all age groups for all cancer types, although the magnitude of improvement varied by age group and cancer type. In most cancers, high-income patients demonstrated the greatest survival improvement from 2014-2018. In colon, rectal, liver, and lung cancers, women had greater improvements in CSS than men, while men had greater improvement in pancreatic cancer. Finally, comparing 2014-2018 to 2004-2008, black patients showed an increase in Stage I diagnoses across all cancer types and a decrease in Stage IV diagnoses for several cancer types, potentially contributing to survival improvement.

The study's findings confirm persistent racial and socioeconomic disparities in cancer survival, even with overall improvements in survival across all groups. The continued lower CSS for Black patients in breast and prostate cancer, despite substantial improvements over time, highlights the need for targeted interventions to address the underlying causes of these persistent inequalities. Differences in access to quality healthcare, disparities in treatment aggressiveness, and socioeconomic factors likely contribute to these disparities. The observation of improved survival linked to earlier stage diagnosis for Black patients suggests that strategies focusing on early detection and access to preventative care could substantially impact these disparities. While the study highlights significant improvements in survival across several demographic groups, the persistent racial disparities warrant further investigation.

This study demonstrates persistent racial and socioeconomic disparities in cancer survival, despite improvements seen across all groups. The fact that black patients in some cancers show improved survival rates but still lag behind white patients highlights the need for targeted research into the underlying factors and the development of specific interventions. Future research should focus on identifying and addressing these underlying causes to achieve true health equity in cancer care. Further investigation into biological factors, access to care, and socioeconomic determinants are essential.

This study utilized hospital-based SEER data, potentially excluding patients treated in private clinics or other settings. Missing data and potential inaccuracies in income (measured at the area level) may introduce bias. Early censoring could affect survival estimates, and the lack of molecular data limited the investigation of receptor status and its interaction with race/ethnicity. The use of area-based measures of income could also introduce limitations and inaccuracies.