Introduction: Chronic Myeloid Leukemia (CML) is a common presentation of leukemia in the United States, representing 15-20% of all newly diagnosed leukemias in adults. Once considered a disease with a grave prognosis, the widespread availability of imatinib, after approval in 2001, improved the estimated 5-year survival of CML to 77%, with an estimate of 90% when patients reliably take tyrosine kinase inhibitors. Due to the increase in survival, patients with CML require longer follow-up than many other hematologic malignancies, allowing for a greater impact of social determinants of health (SDoH) on care. Individual SDoH factors, such as lower socioeconomic status (SES) and race, have already been identified as contributory towards a poorer prognosis in those with CML. However, a systematic assessment of the impact of various SDoH on CML has yet to be described in the current literature. This study compares aggregated SDoH encapsulated by the social vulnerability index, a tool created by the Centers for Disease Control and Prevention (CDC) using census data, to achieve a more complete understanding of the impact of SDoH on prognosis and follow-up in patients with CML.

Methods: All adult patients with CML, as tracked in the Surveillance, Epidemiology, and End Results (SEER) Program, were matched geographically to determine social vulnerability using the Social Vulnerability Index (SVI) at the time of diagnosis. SVI dynamically weighs 15 SDoH considerations at the census tract or county level into 4 themed domains, including minority status & language, housing type & transportation, SES, and household composition & disability, to provide a percentile rank ranging from 0 (least vulnerable) to 1 (most vulnerable) across the United States. CML patients were grouped into equivalently-sized relative-SVI quintiles based on actual SVI scores ranging from lowest to highest vulnerability. Data was analyzed to assess months survival in CML, distributed in quintiles. Univariate linear regressions were used to determine significance. Similar methodologies and analyses were also conducted for SVI-themed domains.

Results: Overall, 34,784 individuals with CML were identified in SEER, with 21,074 of those individuals included in months survival analysis due to loss to follow-up. All groups were statistically significant (P<0.05) with both overall and SVI domains demonstrating decreasing months survival with increasing social vulnerability (Table 1). For the overall vulnerability there was a loss of 2 months of life comparing the highest vs the lowest socially vulnerable quintiles (Figure 1). For domain analysis, comparing the highest to the lowest vulnerability quintiles, there was a 3 month decrease in survival for SES and minority-language status, a 2 month decrease in survival for household composition and a 1 month decrease in survival in housing & transportation.

Conclusion: All specific social vulnerability domains as well as the aggregate overall SVI score correlated with statistically decreased months survival in high social vulnerability patients with CML compared to those in the lowest social vulnerability quintile. Furthermore, the size of difference noted between high and low vulnerability groups in both categories indicate a persistent role SDoH has in patient mortality and may suggest differing impact across various SDoH domains. Despite the use of standard TKIs for CML, health disparities still impact care leading to poorer prognosis for those of higher social vulnerability. This work provides a body of evidence for the necessity for programs aimed at ameliorating the impact of various SDoH on mortality from CML.

Badawy:Pfizer Inc: Research Funding; Global Blood Therapeutics: Consultancy; Forma Therapeutics: Consultancy; CHIESI Farmaceutici S.p.A: Consultancy; Bristol-Myers Squibb (BMS): Consultancy; Sanofi: Consultancy; Vertex Pharmaceuticals Inc: Consultancy; Bluebird Bio Inc: Consultancy.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution