Social Determinants of Health & Cervical Cancer Outcomes at a Large Academic Institution

Purpose: Social determinants of health (SDoH) are associated with cancer outcomes across many disease sites including breast, lung, and colorectal. Locally advanced cervical cancer (LACC) is a curable disease treated by definitive chemoradiation and brachytherapy (CRT-B). We aim to assess how select SDoH factors impact clinical outcomes in LACC after CRT-B.

Methodology: We performed a retrospective chart review of LACC patients treated with CRT-B from 2013 to 2022 across three metropolitan hospitals within a single academic institution. Inclusion criteria were patients with Stage Ib-IV cervical cancer who completed definitive CRT-B with minimum 3 months follow-up; patients who received initial or adjuvant surgery, or palliative radiation were excluded. We evaluated clinical characteristics and SDoH factors available in the medical record. Cox-proportional hazards modeling was conducted to analyze associations between clinical and SDOH factors with overall survival (OS) and progression free survival (PFS). The Kaplan-Meier method was used to estimate 2-year OS and PFS. Clinical significance defined as a p-value < 0.05.

Results: 150 patients with FIGO Stage I (n=30), II (n=38), III (n=62), and IV (n=20) disease were included. With a median follow up of 26.7 months (22.4- 33.2), 2-year OS and PFS for the entire cohort was 72% (CI 95%: 62.6-79.5%) and 56.8% (47.5-65.1%), respectively. Unemployment (HR 2.11; 1.04-4.26, p=0.038), insurance through Medicare/Government (HR 2.91; 1.26-6.71, p=0.012), FIGO IVA/B disease (HR 2.75; 1.08-7.00, p=0.034), and cardiovascular disease (HR 2.84, 1.25-6.47, p=0.013) were significantly associated with worse OS on univariate analysis. Overall treatment time (OTT) was also correlated with OS (HR 1.01; 1.00-1.02, p=0.017). Age (≥65) trended to have worse OS but was not statistically significant (HR 1.96; 0.92-4.15, p=0.081). PFS was significantly associated with insurance through Medicare/Government (HR 2.19; 1.09-4.42, p=0.028) and FIGO IVA/B disease (HR 2.84; 1.31-6.15, p=0.008), but not with employment or OTT. Race, ethnicity, marital status, histology, and smoking status were not associated with OS nor PFS. On multivariate analysis, insurance through Medicare/Government was significantly associated with PFS (HR 2.22, 1.04-4.75, p=0.04) but not OS. Two-year OS and PFS in unemployed versus employed patients were 59.9% (42.9-73.4%) versus 78.6% (61.2-88.9%) (p=0.041) and 49.9% (34.1-63.9%) versus 59.2% (43.4-71.9%) (p=0.263), respectively. Unemployed patients were more likely to be black (p=0.048), a current smoker (p=0.007), uninsured (p= < 0.001), and older age at diagnosis (p=0.001). Reduced OTT and early FIGO stage favored the employed group, but the difference was not significant.

Conclusions: Our study demonstrated SDoH factors, including employment and insurance status were significantly associated with clinical outcomes in patients treated with CRT-B for LACC. Further investigation into how these SDoH factors affect clinical outcomes in patients with LACC is needed to effectively develop intervention strategies.

References: