Predictors of Osteopenia/Osteoporosis in Children with Acute Lymphoblastic Leukemia.

Background: Loss of bone mineral is a major problem in children with acute lymphoblastic leukemia (ALL), resulting in acute and chronic morbidity. About 30–40% of children with ALL will develop osteopenia/osteoporosis (OP) and about 10–15% will suffer from fractures. Identifying a population at high risk for OP is important to implementing a preventive strategy (e.g. biphosphonate therapy). So far predictors of OP in association with ALL in children are uncertain. Hence we undertook the following study to evaluate predictors of OP in children with ALL treated according to Dana-Farber Cancer Institute protocols.

Aim: To evaluate the relationship between lumbar spine bone mineral density (LS-BMD) Z scores in patients with ALL during maintenance therapy and the variables of age at diagnosis (< 10 vs. ≥ 10-years), risk group [Standard (SR) vs. high-risk (HR)], gender (male vs. female) and LS-BMD at diagnosis.

Methods: Children (≤ 18-years) diagnosed with ALL during the period 1995–2006 who were in first clinical remission, were included in the study. LS-BMD was measured using dual-energy X-ray absorptiometry (DEXA) at the time of diagnosis (n=88) and during the maintenance phase of therapy (n=119). The actual values of LS-BMD were expressed as age and gender matched Z-scores based on local population norms. Regression analyses were used to evaluate the risk of osteopenia, defined as LS-BMD Z score < -1.00, and osteoporosis, defined as LS-BMD Z score < -2.00. We evaluated the effect of age at diagnosis, gender, ALL risk category and LS-BMD at diagnosis on the LS-BMD during maintenance phase of therapy.

Results: Of the 119 patients, 19 (16%) were ≥ 10-years of age, 46 (39%) were girls and 41 (34.5%) had HR ALL. At diagnosis 29 of 88 (33%) patients had osteopenia and 6 (6.8%) had osteoporosis whereas, during maintenance therapy, 47/119 patients (39.5%) had osteopenia and 10 (8.4%) patients had osteoporosis. LS-BMD at diagnosis had a positive linear relationship with LS-BMD during maintenance therapy (Pearson correlation coefficient 0.721, p<0.001). Older children and children with HR ALL had a significantly higher risk of osteopenia compared to younger children (p=0.01) and children with SR ALL (p=0.019). Age and risk category were confounding variables since all children ≥ 10 years were classified as HR ALL. Gender by itself had no significant effect. However, the effect of age on the LS-BMD during maintenance phase was gender-dependent with older girls having lower LS-BMD compared to older boys.

Conclusions: Osteopenia and osteoporosis (OP) are common in children with ALL. Age over 10-years, female gender, HR ALL and lower LS-BMD at diagnosis are predictors of lower LS-BMD during the maintenance phase of therapy. Using these variables it is feasible to develop a predictive model to define the risk of OP during the maintenance phase of therapy. Larger prospective studies will better define this risk.