Incidence and Predictors of Early Treatment-Related Mortality In Pediatric Acute Lymphoblastic Leukemia In Baghdad (Iraq)

Abstract 2132

The overall cure rates for children with acute lymphoblastic leukemia (ALL) treated and managed in high-income countries have reached approximately 80% over the last two decades. Unfortunately, these advances in survival have not fully translated into low-income countries where the survival rates remain significantly lower (<35%). Potential reasons for these different results include higher rates of relapse, a high degree of treatment abandonment, insufficient diagnostic work-up procedures, limited availability of effective drugs and supportive measures, and, consequently, high rates of treatment-related mortality (TRM). We examined the incidence, causes and risk factors for early (<60 days) TRM in pediatric patients (≤15 years) with newly diagnosed ALL managed at the oncology unit of the Children Welfare Teaching Hospital in Baghdad (Iraq), over a 3-year period (2007 – 2009). Data were prospectively collected in Baghdad and analyzed at the GIMEMA Data Center in Rome. From January 2007 to December 2009, a total of 319 children (median age 5.2 years, range 0.3–13.9; 171 males and 148 females) with newly diagnosed ALL were registered; the diagnosis of ALL was confirmed by BM aspirate, according to the FAB classification; patients with L3 morphology (6 cases) were included. The median duration of symptoms prior to diagnosis was 4 weeks, ranging from 1 to 76 weeks. At disease onset, 179 children (56%) presented fever and 30 (9.4%) had hemorrhages; liver and renal functions were impaired in 9/290 (3%) and 21/289 (7.3%) patients with available data. The median Hb level was 7.0 g/dl (range 2.4–14.9), the median WBC count was 16.9 × 10⁹/l (range 0.2–900) and the median platelet count was 35.0 × 10⁹/l (range 0.1–598). CSF was positive in 14/290 children (4.8%). Patients were defined as low (153, 48%), intermediate (127, 40%) or high (33, 10%) risk according to clinical and laboratory parameters (age, hepatosplenomegaly, mediastinal mass, WBC, Hb level, platelet count, CNS and testicular infiltration). Sixteen children were discharged following the parents’ decision (14 before any treatment and 2 after 2 days); 303 children are evaluable for early TRM. Treatment consisted of a modified BFM-95 protocol in the first 31, a modified MRC UKALL-2003 protocol in 266 and the LMB/FAB-96 protocol in the 6 children with ALL-L3 morphology. Up to September 25, 2008 all trials did not include the 7-day steroid pre-phase that was introduced thereafter. A total of 249 children (82%) achieved a complete response in the first 60 days of treatment. The cumulative incidence of early TRM was 16% (48/303); it significantly decreased throughout the study period (2007: 21/88, 24%; 2008: 15/98, 15%; 2009: 12/117, 10% p 0.009). Several variables (sex, age, symptoms duration, hepatosplenomegaly, Hb level, WBC count, platelet count, bleeding, fever, impaired liver and renal function, CNS positivity, risk group, steroid pre-phase and induction complications) were examined as potential predictors of TRM. In univariate analysis, the occurrence of induction complications significantly increased the early TRM: hemorrhage 26% vs 11% p 0.001; infection 18% vs 2% p 0.005. A highly significant favorable impact on early TRM was represented by the 7-day steroid pre-phase; 36/169 children (21%) who did not receive the pre-phase died within the first 60 days of treatment compared to 12/134 children (9%) who underwent the steroid pre-phase (p 0.003). When the steroid pre-phase was placed in multivariable models with each of induction complications or other clinical parameters, it remained an independent predictor of TRM.

Our experience confirms that a protocol-based care of children with ALL has to include the prednisone pre-phase that in low-income countries may contribute to a better risk definition and also to a significant reduction of early TRM.