Obesity is associated with poor prognosis in ph-like acute lymphoblastic leukemia

Abstract

Abstract Philadelphia-like (Ph-like) B-cell Acute Lymphoblastic Leukemia (B-ALL) is a subset of high-risk (HR) ALL, constituting 5-15% of pediatric and adult non-Down Syndrome (DS) B-ALL, and is associated with poor survival. Approximately half of Ph-like B-ALL cases exhibit cytokine receptor-like factor 2 overexpression (CRLF2+). Through a series of consecutive trials studying body composition in ALL, we found a strong association between CRLF2+ B-ALL and obesity. CRLF2+ B-ALL was more prevalent in those presenting with obesity, and the odds of being CRLF2+ increased by 13% for every kilogram of fat mass present at diagnosis (Mittelman et al, Blood, 2021). We also previously reported for HR B-ALL that obesity is associated with chemotherapy resistance (as evidenced by minimal residual disease [MRD]) and a poorer prognosis, with a 30-50% increased risk for poorer event free survival (EFS) and overall survival (OS). However, the impact of obesity on outcomes specifically in patients with Ph-like B-ALL is unknown. Given the poor prognosis of Ph-like B-ALL, the association of a common subtype of Ph-like B-ALL with obesity, and the sustained high prevalence of childhood obesity, this is a critical knowledge gap addressed by this study. We conducted a retrospective study of patients treated for non-DS Ph-like B-ALL from 2006 – 2024, with earlier cases diagnosed retroactively. Data included diagnosis (Ph-like subtype), age, sex, self-reported ethnicity, body mass index (BMI), treatment protocol, and clinical outcomes (MRD, EFS, OS). An event was defined as refractory disease, relapse, or death for EFS, and death for OS. Obesity was defined as BMI ≥95th percentile for age and sex for patients aged 19 years and younger, and a BMI of ≥30 kg/m2 for patients over 20 years old. End of induction (EOI) MRD was evaluated in those with available MRD as a binary endpoint (<0.01% versus ≥0.01%). Survival was estimated using the Kaplan-Meier method, and 95% confidence intervals (CIs) were calculated using Greenwood's method. Cox regression was used to associate EFS and OS with potentially prognostic variables (Ph-like subtype, central nervous system (CNS) 1 or 2 versus CNS 3). A chi-square test was used to compare MRD persistence in patients with and without obesity. Survival time was measured from the time of diagnosis until time of event or last follow-up. All statistical analyses were two-sided, with a Type I error rate of 0.05 used to determine significance. This study was exempted per the Institutional Review Board (IRB). Of 80 patients with Ph-like ALL, 66 patients (83%) were included (excluded for DS-ALL [n=8], insufficient treatment/outcome data [n=6]). Of the 66, 47 (71%) were male, 54 (82%) were of Hispanic/Latinx ethnicity, and 40 (61%) had obesity. Ages ranged from 2 to 20 years (median: 16.2 years). The majority of cases were CRLF2+ (either IgH:CRLF2 or P2RY8:CRLF2 fusions, n=57/66, 87%), with the remainder harboring ABL-class fusions, PDGFRB fusions, CSF1R-fusions, or JAK2 mutations. Twelve (18%) patients received a targeted therapy (tyrosine kinase inhibitor or JAK inhibitor). There was no significant difference in MRD persistence in those with versus without obesity (65% vs 63%, p=0.88). On the multivariable EFS analysis, we found evidence for increased risk of an event among patients presenting with versus without obesity (hazard ratio [HR] 2.03; 95% CI 0.942-4.395, p=0.071). In patients who achieved EOI MRD negativity, those with obesity had a significantly lower EFS (p=0.014) than those without obesity (3-year EFS: 59% [95% CI: 31.3-87.3%] versus 100%); however, patients with persistent EOI MRD had poorer EFS regardless of obesity (39% vs 34%, p=0.81). There was no statistically significant difference for OS in patients with versus without obesity overall and regardless of MRD status. Obesity negatively impacted EFS in pediatric patients with Ph-like ALL, predominantly in those with EOI MRD negativity, leading to poor outcomes in this group with expected favorable prognoses. Obesity did not impact OS, indicating that salvage therapies remained successful in achieving sustained remissions for those with obesity failed by frontline treatment, though increased morbidity associated with relapse therapy must be considered. These findings add to the growing literature implicating obesity as a poor prognostic factor in higher risk acute leukemias. Further studies will explore obesity's impact on molecular drivers of Ph-like ALL.