LDL Discordance in Pediatric Survivors of Acute Lymphoblastic Leukemia

Background: Survivors of childhood acute lymphoblastic leukemia (ALL) have a four-fold excess risk of mortality secondary to cardiovascular events compared to the general population. While it is well-known that LDL has a role in atherogenesis and subsequent cardiac events, recent studies consistently show that LDL particle number (LDLp) is a stronger, independent predictor of coronary heart disease in adults compared to LDL-cholesterol (LDLc), especially when the values are discordant. There is emerging evidence that LDL discordance also occurs in children. We studied this phenomenon in pediatric ALL survivors and identify associated risk factors.

Methods: Complete data were available from 64 patients enrolled in the Phoenix Children’s Hospital pediatric cancer survivorship registry. Patients were 12.9±3.5 years of age (range: 7.7-22.7 years) with a mean time off therapy of 5.5±2.6 years (range 2.3-11.4 years). LDLc and LDLp were assessed by NMR spectroscopy. Patients were assessed by a registered dietician for calcium, fat, and fruit intake as well as physical activity. Ideal LDLc and LDLp were defined as <160 mg/dL and <1600 nmol/L, respectively, based on American Heart Association guidelines for people with standard cardiac risk. Discordance was classified based on ideal vs. not-ideal status for each parameter.

Results: Mean BMI was 22.7±5.7 kg/m² (range 14-38 kg/m²); mean BMI percentile was 73.1±26.1% (0.9-99.5%) with 42% of patients considered overweight or obese at the time of evaluation. Mean LDLc and LDLp were 79±25.5 mg/dL (29-164 nmol/L) and 1124±517 nmol/L (411-2837 nmol/L), respectively, with 89% (n=56) of survivors exhibiting ideal LDLc and 47% (n=30) of survivors exhibiting ideal LDLp. In a subgroup of patients with ideal LDLc, 40.6% (n=26) exhibited not-ideal LDLp and were classified as discordant. Regression analysis showed that after adjusting for age and gender, BMI (p<0.00001) and LDLc (p<0.00001) were significant independent predictors of LDLp while lifestyle factors and time off therapy were not significant predictors of LDLp. When patients were grouped as concordant (ideal LDLc and LDLp, n=30) or discordant (ideal LDLc and elevated LDLp, n=25), BMI was 17.6% higher in the discordant group (25.0±5.4 kg/m² vs. 20.6±5.0 kg/m², p=0.003).

Conclusion: In this population of pediatric ALL survivors, LDLc measurements may not provide a complete assessment of cardiovascular disease risk. Given that LDLp is an early predictor of cardiovascular outcomes, conventional lipid testing may underestimate the true cardiovascular disease risk in ALL survivors. Survivors may benefit from early, expanded screening and targeted intervention. Further studies should focus on the broader pathophysiology of cardiovascular disease in pediatric ALL survivors.