Cranial Irradiation for Pediatric T-Lineage Acute Lymphoblastic Leukemia: A Systematic Review and Meta-Analysis.

There is currently a heterogeneous approach to the use of cranial irradiation (CRT) for central nervous system (CNS) directed therapy for T-lineage acute lymphoblastic leukemia (T-ALL) by pediatric cooperative groups. We sought to explore the association of CRT and event-free survival (EFS) in children with T-ALL by means of a systematic review and meta-analysis.

We searched MEDLINE from 1949 to 2012 without geographic restrictions, to identify English language randomized trials or cohort studies reporting EFS in children with T-ALL who received CNS directed therapy. For studies to be eligible a CRT strategy needed to be described (intrathecal chemotherapy +/− cranial irradiation) and at least 3-year EFS (or longer) needed to be reported. Studies need to include at least 10 pediatric subjects with T-ALL; there was not a minimum proportion of T-ALL subjects needed to meet eligibility criteria. The following study characteristics were extracted for each study: eligibility criteria, patient number, CRT strategy, intrathecal chemotherapy administered and number of doses, steroid administered in induction, cumulative doses of high-dose methotrexate, asparaginase, and anthracyclines, definition of event-free survival, median follow-up, and event-free survival. We conducted subgroup analyses (random effects) and random effects meta-regressions to explore associations between EFS and the following study-level factors: (i) CRT strategy, categorized as cranial irradiation for all subjects, cranial irradiation administered in a risk-directed fashion (often stratified by age and WBC count at diagnosis), cranial irradiation for CNS positive patients only, and omission of CRT for all patients; (ii) enrollment year (continuous); (iii) intrathecal chemotherapy (methotrexate vs. triple intrathecal chemotherapy); (iv) maximum number of intrathecal chemotherapy dose (<10 vs. 10–19 vs. > 20); (v) high-dose methotrexate (dose > 1 gram/m2) present or absent; (vi) high cumulative dose of asparaginase (> 400,000 IU/m²) present or absent, (vii) high cumulative dose of anthracyclines (doxorubicin plus daunorubicin total > 300 mg/m²)present or absent; (viii) induction steroid (prednisone vs. dexamethasone).

The search returned 2383 abstracts, 491 of which were reviewed in full text. Eligible were 59 articles (5726 T-ALL patients enrolled between 1973 and 2005). The overall 3-year EFS was 62.1% (95% CI: 58.9% to 65.3%). There was significant heterogeneity among the treatment studies (I2̂ = 80%, p < 0.001). An improvement in EFS was significantly associated with the year study enrollment began (p< 0.001); based on a meta-regression model of EFS versus year of enrollment start, the average EFS improved from 44% for studies that began in 1980 to 71% for studies that began in 2000.

The EFS was significantly different across the 4 CRT categories (omnibus p-value=0.02): CRT to all patients (EFS: 63%, 95% CI: 58% to 67%) risk-directed CRT (EFS: 59%, 95% CI: 53% to 65%), CRT for CNS positive patients only (EFS: 55%, 95% CI: 41% to 68%), CRT omitted for all patients (EFS: 74%, 95% CI: 67% to 80%). The association of EFS and CRT strategies remained in the same direction after adjusting for year of enrollment.

The following factors were also associated with an increased odds of EFS on univariate analysis: the administration of 10–19 or > 20 doses of intrathecal chemotherapy (OR: 1.84, 95% CI: 1.32 to 2.58 and OR: 1.95, 95% CI: 1.31 to 2.94, respectively), the administration of high dose methotrexate (OR: 1.43, 95% CI: 1.06 to 1.93), and the administration of more than 400,000 IU/m2 of asparaginase (OR: 1.79, 95% CI: 1.27 to 2.51). However after adjusting for the year of enrollment, only high doses of asparaginase remained significantly associated with EFS.

Our systematic review and meta-analysis found that CRT strategy is associated with EFS for children with T-ALL. Studies that omitted CRT were associated with superior EFS; however, these studies were more recent, and there was an improvement in EFS with time. The administration of more than 400,000 IU/m² of asparaginase was also associated with lower relapse rates. Although these noncausal associations are congruent with the notion that CRT may not offer an improvement in survival for children with T-ALL given current chemotherapeutic options, the existing evidence-based is not sufficient to draw such a conclusion.

Off Label Use: asparaginase is a a drug used to treat acute lymphoblastic leukemia.