Age, Sex, Timing of Treatment Response Predict Avascular Necrosis in Children with ALL

Avascular necrosis (AVN), a disorder resulting from a temporary or permanent loss of blood supply to the bone, is one of the most serious late complications of therapy in children with acute lymphocytic leukemia (ALL). Currently, there is no consensus on how osteonecrosis in ALL develops.

In a study presented at the 2015 ASH Meeting on Hematologic Malignancies, Ksenia Romanova, from the Institute of Oncology, Radiology, and Nuclear Medicine at the Federal Research Center of Pediatric Hematology, Oncology, and Immunology in Moscow, Russia, and colleagues identified risk factors associated with development of AVN in children with ALL.

“This condition leads to disability, requires surgery to conduct joint replacement, and lengthens the time of therapy and increases its cost,” Dr. Romanova and co-authors wrote. Strategies to predict risk and guide appropriate treatment are needed.

The researchers analyzed data from children with ALL who had received therapy within the multicenter trial protocols Moscow-Berlin 2002 (MB-2002) and Moscow-Berlin 2008 (MB-2008) from April 2002 to November 2014. Overall, 5,205 patients were included: 1,544 from MB-2002 and 3,661 from MB-2008.

After adjusting for sex, age, type of glucocorticoid therapy during induction (for MB-2002), therapy with and without PEG-asparaginase (for MB-2008), and the different post-induction therapy arms within each protocol, the authors found that incidence of AVN was “lower than expected from data reported by other groups: 1.3 percent in the MB-2002 group and 0.9 percent in the MB-2008 group (p=0.21).

Despite an overall low incidence, there were several risk factors associated with the development of AVN: age over 10 years, high body mass index at the beginning of treatment, and lack of response on the 15th day of treatment (TABLE).

While there was no significant difference by gender, multivariate analysis revealed a predominance of female patients in the group of older children with AVN (6.1% for girls, 3.1% for boys; p=0.03).

Dr. Romanova and colleagues suggest that “the [lower-than-expected incidence of AVN] could be explained by the lower cumulative dose of glucocorticoids and the avoidance of high-dose chemotherapy (no cyclophosphamide and use of high-dose methotrexate in a limited group of patients).”

Patients treated with methylprednisone 120 mg/m2 during induction had a higher incidence than those treated with methylprednisone 60 mg/m2 or with dexamethasone 6 mg/m2: 4.0 percent versus 1.4 percent and 0.9 percent, respectively (p=0.017).

Dose of methotrexate during consolidation, as well, led to changes in AVN: 2,000 mg/m2 methotrexate was associated with higher incidence than 30 mg/m2 (3.5% vs. 1.7% for MB-2002; 1.1% vs. 0.6% for MB-2008). These associations were not clinically significant, though.

Different asparaginase strategies failed to significantly affect the incidence of AVN.


Romanova K, Roumiantseva J, Lagoyko S, et al. Low risk of avascular necroses in children and adolescents with acute lymphoblastic leukemia treated with reduced intensity Moscow-Berlin regimens. Abstract #4. Presented at the 2014 ASH Meeting on Hematologic Malignancies; September 19, 2015; Chicago, IL.

TABLE. Incidence Rates of Avascular Necroses
  Incidence of AVN p Value
Male 0.8% 0.19
Female 1.2%
>10 years 4.3% 0.001
≤10 years 0.8%
Body mass index at beginning of treatment
High 3.1% 0.0001 for all
Normal 1.6%
Low 0.3%
Response to treatment on day 15
>10% blast cells in the bone marrow 1.8% 0.057
≤10% blast cells in the bone marrow 0.9%