Social Risk Phenotypes Are Strongly Associated With Hepatocellular Carcinoma Outcomes: A Statewide Cohort Study Using Latent Class Analysis.

[INTRODUCTION] Disparities in hepatocellular carcinoma (HCC) outcomes are shaped by intersecting social determinants of health. We hypothesized that patients experience distinct combinations of socioeconomic barriers that cluster into social risk phenotypes associated with differences in diagnosis, treatment, and survival.

[METHODS] We analyzed data from 4,877 adults diagnosed with HCC in the Indiana State Cancer Registry (2009-2020). Latent class analysis was performed using sex, race, insurance, marital status, occupation, neighborhood Social Deprivation Index, and distances to screening and Indiana University Hospital. Outcomes included early-stage diagnosis, receipt of curative therapy, and 2-year mortality.

[RESULTS] Among 4,877 patients, 15.8% were non-White and 24.7% were female. Latent class analysis identified 6 distinct risk classes: (i) minimal barriers, (ii) publicly insured-married women, (iii) publicly insured-unpartnered men, (iv) rural and geographically distant, (v) structurally marginalized, and (vi) unseen and uninsured. Class 1 had the most favorable characteristics (83.7% private insurance, 16.8% professional occupation) and best outcomes: 55.4% early-stage diagnosis, 24.4% curative therapy, and 55.4% 2-year mortality. All other classes had significantly worse outcomes. Compared with class 1, patients in class 6 had the lowest early-stage diagnosis (39.7%) and curative therapy (10.5%) and highest 2-year mortality (83.6%; odds ratio 4.12, 95% confidence interval 3.06-5.54, P < 0.001). Classes 4 and 5, reflecting rural and racially marginalized groups, also had significantly lower odds of early diagnosis and treatment.

[DISCUSSION] Social risk phenotypes based on intersecting social determinants of health were strongly associated with HCC outcomes and may inform future risk-based intervention strategies.