A critical review of the prognostic value of the nutritional status at diagnosis in the outcome of therapy of children with acute lymphoblastic leukemia.

The impact of undernutrition in the outcome of treatment of children with ALL has been analyzed by several authors who have highlighted undernutrition as another relevant prognostic factor in children with acute lymphoblastic leukemia (ALL). There are, however, some papers which have not confirmed the prognostic value of malnutrition at diagnosis in children with ALL. Overall, data from 1,123 children with ALL worldwide support the concept of malnutrition at diagnosis being useful as a prognostic factor, whereas data from 1,271 children fail to support this concept. We here critically analyze the information of these publications referring to a total of 2,394 children with ALL. Detailed information was available only from 500 of the 2,394 patients, stemming from six publications; of these individuals, at diagnosis, 376 were well nourished and 124 were malnourished. In this subset of patients, the analysis of the data shows that the 5-year (or longer) overall survival of undernourished children (UNC) was 26%, whereas that of well-nourished children (WNC) was 59% (p < 0.001); along the same line the relative risk of dying during this period was 1.8 times higher for UNC than WNC (p < 0.01; Interval of Confidence [IC] 95%: 1.72-1.88). On the other hand, the censoring time of these 500 children is different: 293 were censored at 5 years after diagnosis, whereas 207 where censored at 8-10 years: The overall survival for each of these periods was also different: 36% versus 2% for UNC (p < 0.001), and 56% versus 63% (p > 0.10) for WNC, data which support the concept that the differences in survival are more apparent in UNC if the period of observation is extended. These data suggest that undernutrition at diagnosis by itself and without interacting with other variables, may be a significant prognostic factor in the long-term outcome of treatment of pediatric patients with ALL. After identifying these variables as important, imaginative approaches to the treatment of cancer in childhood in the years ahead may lead into the improvement of the results of these treatments.