Impact of geographic distance and residential setting on overall survival in patients with Acute Myeloid Leukemia Free

Background:

The treatment of acute myeloid leukemia (AML), particularly in the initial (induction) phase, is typically done in large-volume leukemia treatment centers, often academic hospitals due to the immensity of resources required. While several studies have investigated the association between geographic distance to the treatment center and overall survival (OS) in patients with AML, the findings have been inconclusive or demonstrated no significant association. These mixed results may be attributed to the presence of multiple confounding variables. In this study, we aim to evaluate the impact of distance to the University of Tennessee Medical Center (UTMC) in Knoxville on OS in AML patients, while also examining the influence of patients' residential settings, categorized as rural or urban, based on the Rural-Urban Commuting Area (RUCA) score.

Method:

We conducted a single-center retrospective study including 68 newly diagnosed AML patients, selected from an initial cohort of 300 patients between January 2022 and December 2024, after excluding cases of secondary AML, acute promyelocytic leukemia, and patients not treated at our institution. Distance to the UTMC treatment center and RUCA score were calculated based on postal codes. Associations between distance to the treatment center, RUCA score, and overall survival were first examined using ANOVA. Recognizing that age and disease risk stratification (using the European LeukemiaNet (ELN) risk stratification) could confound these relationships, we subsequently utilized a Cox proportional hazards model to adjust for these factors and assess the independent impact of distance and RUCA score on survival outcomes.

Result:

The mean distance to the treatment center was 33 miles, and the mean overall survival was 14.5 months. Distance to the treatment center was stratified into four categories: ≤13.3 miles, 13.4–29.0 miles, 29.1–45.5 miles, and >45.5 miles. RUCA (Rural-Urban Commuting Area) scores were grouped into four categories based on degree of urbanization: Group 1 (RUCA 1, metropolitan), Group 2 (RUCA 2–3, suburban), Group 3 (RUCA 4–6, small town), and Group 4 (RUCA 7–10, rural). ANOVA revealed no statistically significant differences in overall survival across distance groups (p = 0.90), with median survival times of 14.29 ± 10.52 months for ≤13.3 miles, 14.33 ± 12.95 months for 13.4–29.0 miles, 14.63 ± 9.33 months for 29.1–45.5 miles, and 14.64 ± 13.36 months for >45.5 miles. Similarly, no significant differences in survival were observed across RUCA groups (p = 0.33), with median survival times of 12.63 ± 10.01 months for Group 1 (metropolitan), 13.08 ± 9.52 months for Group 2 (suburban), 18.45 ± 15.39 months for Group 3 (small town), and 18.01 ± 12.98 months for Group 4 (rural). In the Cox proportional hazards model, none of the covariates (age, ELN risk category, distance, or RUCA score) were independently associated with survival (p > 0.1). However, the global likelihood ratio test indicated a significant overall model fit (p = 0.02).

Conclusion:

Compared to previous studies examining the relationship between treatment center proximity and AML survival, our findings were largely consistent. However, our analysis uniquely incorporated both geographic distance and residential setting, as measured by RUCA score. While the RUCA-based analysis yielded a lower p-value than distance, it did not reach statistical significance. Notably, the overall likelihood ratio test for the Cox proportional hazards model was significant, suggesting a potential joint effect of age, ELN risk category, distance, and RUCA score on survival outcomes. Given the relatively small sample size and limited follow-up period in our study population, this finding suggests that unmeasured or complex interactions among these variables may influence prognosis in patients with AML. Future studies with larger cohorts, extended observation periods, and integrated analyses of these factors are warranted to clarify their potential contributions to survival outcomes.