Hydroxyurea Use Is Associated with Avascular Necrosis of the Femoral Head among Children with Sickle Cell Disease

Hydroxyurea therapy is associated with reduced morbidity among patients with sickle cell disease (SCD). Avascular necrosis of the femoral head (AVN) is one potentially debilitating complication of SCD. In this study, we examined the relationship between hydroxyurea use and the prevalence of AVN among children with SCD. We performed a retrospective chart review of 202 children with SCD, aged 10–21 years, followed in the pediatric hematology program at the Children’s Hospital at Montefiore (Bronx, NY) between July 2007 and 2008. Abstracted data included age, ethnicity, SCD genotype, frequency of hospitalization, hip radiograph results, laboratory data and hydroxyurea use. Hip radiographs were performed prospectively as part of SCD health maintenance from 2005–2008. Forty-four patients were excluded because they did not have a screening hip radiograph. Descriptive statistics were calculated for independent variables. T-tests and chi-square tests were used to compare clinical and demographic characteristics of children with and without AVN. Multivariate logistic regressions were used to estimate the odds ratio of having AVN among SCD patients. Our final sample consisted of 158 patients whose demographic characteristics are listed in Table 1. The prevalence of AVN was 16.5% (n=26). Of the clinical variables analyzed, we identified significant associations between the presence of AVN and hydroxyurea use (p=.005), as well as older age (p=.013) (Table 1.) Children with AVN had significantly lower mean lactic dehydrogenase levels (LDH) (p=.04) and higher mean corpuscular volumes (MCV) (p=.012). (Table 2.) After controlling for gender, ethnicity, sickle cell genotype, and frequency of hospitalizations, age was also found to be associated with AVN (OR 1.15, 95% confidence interval (CI): 1.01,1.31, p=0.033). SCD patients on hydroxyurea had higher odds of having AVN compared to non-users (OR 3.51, 95% CI: 1.31, 9.38, p= 0.013). Laboratory values (MCV, Hemoglobin, LDH and Hematocrit) had a high degree of collinearity and were removed from the final model. In summary, the prevalence of AVN in our sample was 16.5%. This is substantially higher than the prevalence of approximately 6% reported by the Cooperative Study of Sickle Cell Disease for comparative age groups in a prospective study¹. SCD patients exposed to hydroxyurea were three times more likely to have AVN than those not exposed to this drug. Vaso-occlusive pain crisis is a recognized risk factor for AVN, thus we could expect a higher rate of AVN among patients on hydroxyurea. However, the odds ratio of 3.5 is unexpectedly high and warrants further investigation into the role of hydroxyurea as a risk factor for AVN. Nonetheless, these preliminary results suggest that more stringent screening regimens for AVN may be indicated among this subset of patients.

Table 1. Clinical characteristics of patients with and without avn

Table 2. Mean Laboratory Values for Patients With And Without AVN