Outcomes for acute lymphoblastic leukemia (ALL) reported by high-income countries have historically not been reproducible in low- and middle-income countries (LMIC). In Mexico, 5-year event-free survival rate (EFS) for ALL is close to 50% and reasons for the inability to reproduce HIC results include excess of treatment-related mortality (TRM) and high relapse rates (RR). Although higher intensity is bound to lead to lower RR, the increased TRM seen in LMIC abrogates this benefit, leading to poor outcomes.


In April 2016, the group Mexico in Alliance with St. Jude (MAS) was created with the aim to increase survival of children with cancer in Mexico via multi-site collaboration. Work began with ALL, considering its high incidence (89.5 cases per million per year). A retrospective review, including 207 patients diagnosed in 2015 at the 8 member centers was initially performed. Analysis of early therapy results in November 2016 (done to document baseline clinical characteristics and major contributors to early treatment failure) showed, 82% of patients classified as high-risk for reasons including: Leukocytosis >50,000 (27%), age >10 years (36%), T-cell immunophenotype (8%), CNS3 (4%), hypodiploidy (3%); mediastinal mass (4%), testicular infiltration (3%), poor response to prednisone (14%), prior treatment (8%), MLL (1%); iAmp21 (1%), and BCR-ABL (1%). Molecular studies and MRD were not send in up to a third of patients. Among MRD samples sent, 28% were interpreted as high, suggesting difficulty with distinction between leukemic blasts and hematogones. Excess of high-risk patients was noted, but accounted for from use of DNA index as independent predictor for high-risk disease, prior treatment, continuation of therapy after treatment abandonment, lack of MRD validation, and missing of a prognostic factor (due to lack of access to molecular studies, among others). TRM during induction was 12%, primarily as a result from infection, induction failure 2%, and a treatment abandonment rate 4%. At time of analysis, 75% were alive and under treatment, with RR of 8%. Documented survival was therefore already considerably lower than expected for the early follow-up. Unexpectedly high prevalence of classic adverse risk factors was not identified. Early survival was most again affected by high TRM rate during induction and early treatment.

Results: Considering these data, three main lines of action were identified: a) Modernization of risk assignment across the 8 sites through collaborative regionalization of specialized diagnostic techniques, which allows homogeneous access to a pre-determined molecular biology panel and two MRD time points, and classification in 3 rather than 2 risk-groups; b) Reduction of TRM via implementation of a supportive therapies portfolio inclusive of 4 validated strategies and application of treatment de-intensification when feasible; and c) Establishment of a data management system to facilitate documentation of outcomes.


The developed protocol standardizes the introduction of three risk groups: Favorable-risk (FR), Intermediate-risk (IR) and High-risk (HR) based on clinical, genetic and biologic criteria. All patients share a de-intensified 4-drug remission induction therapy, undergo early response evaluation using MRD by flow cytometry, proceed to intensification unless induction failure is documented and receive final risk assignment after a second MRD check at the end of intensification. Intensification occurs with intermediate (FR) or high-dose (IR and HR) methotrexate in association with E. coli Asparaginase (ASP), considering the protective effect from methotrexate toxicity of such association described by Capizzi, a lower rate of allergic reactions to ASP when the time without ASP is reduced, and lack of access to Erwinia ASP or PEG. CNS prophylaxis with radiation therapy is completely omitted and replaced by lumbar punctures. FR patients receive one reinduction without anthracyclines, IR patients one reinduction with anthracyclines, and HR patients two reinductions. Continuation therapy continues for a total of two years, with three intensity levels defined based on risk group.


Early adoption of new diagnostic technologies and of opportunities for de-intensification are crucial in LMIC in order to identify patients with favorable risk disease and reduce the burden of early mortality.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.