The Impact of Race and Socioeconomic Status (SES) on Survival in Non-Hodgkin’s Lymphoma (NHL): A Population-Based Study from the California Cancer Registry (CCR).

Background: Despite the existence of well-established clinical prognostic indices for diffuse large B-cell lymphoma (DLBCL) and follicular lymphoma (FL), the 2 most common subtypes of NHL, significant heterogeneity in survival remains even within prognostic groups. SES has not previously been evaluated as a prognostic factor for NHL, particularly using an unselected patient population. We therefore used the large multiethnic CCR to examine NHL survival according to histology, race/ethnicity and neighborhood SES.

Methods: NHL pts were identified from the CCR for the period 1988–97 using standardized ICD-0-3 classifications (morphology codes 9690–99 for FL & 9680–84 for DLBCL). Patients with evidence of HIV/AIDS were excluded. Neighborhood SES was assigned based on the address of residence at diagnosis, according to 1990 US Census Bureau census block group (each block contains about 1500 residents), and is based upon principal components of 7 indicator variables of SES (education level, proportion with blue collar job, proportion unemployed, median household income, proportion below 200% of poverty line, median rent & median home value). Using this index, we assigned each patient into an SES quintile (SES-1 lowest, SES-5 highest) based on the statewide distribution of neighborhood SES. We computed 5-year relative survival ± standard error (SE) with SEER*Stat software using customized race and SES-specific life tables based on US Census Bureau estimates for California residents.

Results: Unselected DLBCL pts [n=13,604; comprising 73% non-Hispanic White (W), 4% Black (B), 14% Hispanic (H), 8% Asian/Pacific Islander (A)] & FL pts (n=7372; 82% W, 3% B, 11% H, 4% A) were identified. The overall 5 yr relative survival for DLBCL was 45.5% (SE 0.5) and for FL was 71.1% (SE 0.6); females had a better survival than males for DLBCL [F 48.3% (SE 0.7) vs. M 42.3% (SE 0.6)] but not for FL. Lower SES was associated with inferior survival for both DLBCL & FL (Table 1). Within SES groups there were not significant racial/ethnic differences in survival. However, in DLBCL, B (34%) and H (33%) pts were proportionally more likely to be in SES-1 than W (9%) or A pts (15%), and less likely to be in SES-5 (9% B, 10% H) than W (27%) or A (22%) pts. Similarly, in FL, B (39%) & H (26%) pts were also more likely than W (8%) & A (10%) pts to be in SES-1 and less likely to be in SES-5 (W-29%, A-32%, B-10%, H-11%).

Conclusion: SES predicts survival in DLBCL and FL. Race/ethnicity differences in SES distribution are apparent in NHL patients, but when stratified by SES, race/ethnicity does not appear to predict for significant differences in NHL survival.