Cancer Outcome Disparities Not Fully Explained by Race, Socioeconomic Factors

Enrollment in clinical trials failed to overcome factors leading to worse survival for Hispanic children with high-risk neuroblastoma, a large retrospective analysis showed.

The Hispanic subgroup had a 5-year overall survival (OS) rate of 47% as compared with over 60% for whites and non-Hispanic Blacks. A similar disparity was seen for children in the racial/ethnic category of non-Hispanic other.

Children living in poverty also had worse OS, but socioeconomic differences did not fully explain the lower OS in Hispanic children, said Puja J. Umaretiya, MD, of Dana-Farber Cancer Institute and Boston Children’s Hospital, during a press briefing that preceded the American Society of Clinical Oncology (ASCO) annual meeting.

“Our data showed that Hispanic children with high-risk neuroblastoma have inferior overall survival compared with other children, and this is despite having access to the same clinical trials and receiving the same uniform plan of treatment and after controlling for disease-associated factors,” said Umaretiya. “We acknowledge that describing disparities is simply not enough. We must strive to identify mechanisms that cause inequitable outcomes.”

“In this cohort, we begin to do so by looking at things like stopping the trial early, delays in therapy, and increased rates of relapse, none of which seem to explain the survival differences that we see by race and ethnicity,” she continued. “Proxy poverty exposures did not remain significantly associated with survival after adjusting for other factors, such as race, ethnicity, and disease-associated factors, acknowledging the intersectionality of these three factors in the United States due to structural racism.”

Rather than offer possible explanations for disparate outcomes, the study raised new questions for future study, said ASCO President Everett Vokes, MD, of the University of Chicago, who moderated the press briefing.

“This is really important work,” said Vokes. “We do encourage patients to enroll in clinical trials, and that actually happened here. This was a cohort of patients where treatment was regimented through these clinical trials, and it should have led to similar outcomes, but it didn’t. That, of course, leaves questions open.”

“We see discrepancies in overall survival … of about 8-10% at longer terms of follow-up, and they seem to correlate with socioeconomic factors and race, but the specific causes of death, the increased death in those groups, are not elucidated yet. I think that will be important work moving forward,” he added.

The presentation was one of four in the opening press briefing, which emphasized the theme for the 2022 annual meeting, “Advancing Equitable Care Through Innovation.” Umaretiya presented findings from a review of outcomes in three Children’s Oncology Group clinical trials in newly diagnosed high-risk neuroblastoma. The trials included a cumulative total of 696 pediatric patients with a racial/ethnic breakdown of 69% non-Hispanic white, 16% non-Hispanic Black, 11% Hispanic, and 4% non-Hispanic other.

Previous studies showed that Black children with neuroblastoma had an increased rate of late relapse and that children covered by public insurance had an increased relapse rate and mortality risk when treated with immunotherapy. Explanations for the disparities remained unclear, said Umaretiya. Clinical trials offer an ideal setting to study disparities because children received highly standardized care that would be expected to produce similar outcomes in all patients.

The analysis had two primary objectives: identify associations between race, ethnicity, and socioeconomic status and survival; and explore potential mechanisms for worse survival, specifically early trial discontinuation, delays in induction therapy, and relapse.

The investigators compared survival by two different surrogate measures of poverty: coverage by public insurance and living in a zip code where greater than 20% of residents have household income below the federal poverty level. A third of the children enrolled in the trials met household poverty criteria, a fourth met neighborhood (zip code) poverty criteria, and 15% lived in rural areas, another potential source of outcome disparities.

The analysis showed that non-Hispanic Black and non-Hispanic white patients had a 5-year OS of 61-62%, as compared with 47% for Hispanic patients and 50% for non-Hispanic other (P=0.047). The hazard ratio for death was 1.78 for Hispanic patients versus non-Hispanic whites and 1.45 for non-Hispanic other versus non-Hispanic whites (P=0.01). Non-Hispanic Black patients had a mortality risk similar to that of non-Hispanic whites (HR 0.98).

Children covered by public insurance had a 5-year OS rate of 53% versus 63% among patients with private insurance (P=0.036). Living in a high-poverty neighborhood was associated with a 5-year OS of 54% versus 62% among patients living in more affluent areas (P=0.050). Children living in more rural areas did not have worse survival.

Examination of potential mechanisms or explanations for inferior 5-year OS revealed no significant association with delays in therapy by race/ethnicity (median 167-169 days, P=0.91) or early discontinuation in the trials (32-43% P=0.81 and P=0.97). Hispanic patients exhibited a trend toward higher relapse rate (57% vs 48-49%) and non-Hispanic other patients had a numerically lower relapse rate (30%), but relapse rate did not differ significantly across subgroups (P=0.11).