Elsevier

European Journal of Cancer

Volume 46, Issue 18, December 2010, Pages 3280-3286
European Journal of Cancer

Predictors of outcome and methodological issues in children with acute lymphoblastic leukaemia in El Salvador

https://doi.org/10.1016/j.ejca.2010.07.001Get rights and content

Abstract

Background

Most children with cancer live in low-income countries (LICs) where risk factors in paediatric acute lymphoblastic leukaemia (ALL) developed in high-income countries may not apply.

Methods

We describe predictors of survival for children in El Salvador with ALL. We included patients <16 years diagnosed with ALL between January 2001 and July 2007 treated with the El Salvador–Guatemala–Honduras II protocol. Demographic, disease-related, socioeconomic and nutritional variables were examined as potential predictors of event-free survival (EFS) and overall survival (OS).

Results

260/443 patients (58.7%) were classified as standard risk. Standard- and high-risk 5-year EFS were 56.3 ± 4.5% and 48.6 ± 5.5%; 5-year OS were 77.7 ± 3.8% and 61.9 ± 5.8%, respectively. Among standard-risk children, socioeconomic variables such as higher monthly income (hazard ratio [HR] per $100 = 0.84 [95% confidence interval (CI) 0.70–0.99; P = 0.04]) and parental secondary education (HR = 0.49, 95% CI 0.29–0.84; P = 0.01) were associated with better EFS. Among high-risk children, higher initial white blood cell (HR per 10 × 109/L = 1.03, 95% CI 1.02–1.05; P < 0.001) predicted worse EFS; socioeconomic variables were not predictive. The difference in EFS and OS appeared related to overestimating OS secondary to poor follow-up after abandonment/relapse.

Conclusion

Socioeconomic variables predicted worse EFS in standard-risk children while disease-related variables were predictive in high-risk patients. Further studies should delineate pathways through which socioeconomic status affects EFS in order to design effective interventions. EFS should be the primary outcome in LIC studies.

Introduction

In high-income countries (HICs), over 80% of children with acute lymphoblastic leukaemia (ALL) are cured.1, 2 Unfortunately, cure rates are inferior in low-income countries (LICs), where the majority of children with cancer reside.3 Various phenomena may account for this survival gap, including more treatment toxicity, higher rates of relapse and abandonment of therapy in LICs.

Improvement in survival in HICs has been dependent on the identification of clinical and biological predictors of outcome and the development of various risk stratification schemas.4 This in turn has allowed the tailoring of treatment intensity to disease aggressiveness, avoiding both over- and under-treatment of individual patients. The success of this approach has resulted in several traditional adverse prognostic factors losing their predictive power in the context of modern treatment.1, 2

Risk stratification systems developed in HICs may not, however, be appropriate for children with ALL in LICs. Several routine risk factors, such as cytogenetics and minimal residual disease, are not commonly available in low-income centres. In addition, differences in the underlying biology and the external circumstances are likely to require the adaptation of traditional risk strata. Socioeconomic and nutritional factors may also play important roles.

Finding accurate predictors of outcome is essential if LICs are to close the survival gap which currently exists in ALL. The appropriate balancing of treatment intensity to risk of relapse and treatment-related mortality (TRM) is of particular importance in LICs given the significant resource constraints and difficulties in supportive care.5 Despite this, few investigators have systematically examined the risk factors for survival in paediatric ALL in LICs. The primary objectives of this study were therefore to determine predictors of event-free survival (EFS) in children with ALL treated in El Salvador, an LIC. The secondary objectives were to describe overall survival (OS) and the relationship between EFS and OS in this population.

Section snippets

Data source

Data were obtained from two sources. The primary source of data was the Paediatric Oncology Networked Database (POND) (www.pond4kids.org). POND is an online database for paediatric cancer patient information designed to permit users at multiple locations to store and analyse data that include patient demographics, diagnoses, treatments and outcomes in a secure environment with stringent control of access and privacy.6 Although data collection is primarily to assist in patient outcome monitoring

Results

The study sample included 443 children with ALL treated according to the EGH II protocol. Two hundred and sixty (58.7%) were classified as standard risk while the remainder classified as high risk. Demographic characteristics of the patient population can be seen in Table 1.

The 5 year EFS was 56.3 ± 4.5% and 48.6 ± 5.5% for the standard- and high-risk cohorts, respectively. The 5 year OS was 77.7 ± 3.8% and 61.9 ± 5.8% for standard- and high-risk patients (Fig. 1). There were 82 events among

Discussion

This study found that among the children with ALL treated in El Salvador, the 5 year EFS was 56.3 ± 4.5% and 48.6 ± 5.5% for the standard- and high-risk patients, while the 5 year OS was 77.7 ± 3.8% and 61.9 ± 5.8%, respectively. Among the standard-risk cohort, socioeconomic variables predicted EFS, while only disease-related variables did so in the high-risk children.

One striking finding is the wide and sustained gap between EFS and OS in both standard- and high-risk patients. In HICs, this difference

Sources of support

This work was supported in part by the Paediatric Oncology Group of Ontario, the American Lebanese Syrian Associated Charities and the Fundación Ayúdame a Vivir de El Salvador.

Conflict of interest statement

None declared.

Acknowledgements

L.S. is supported by the Canadian Childhood Clinician Scientist Training Program, a training program of the Canadian Institutes of Health Research. This work was supported in part by the Paediatric Oncology Group of Ontario, the American Lebanese Syrian Associated Charities and the Fundación Ayúdame a Vivir de El Salvador. Funding sources had no role in the writing of the manuscript or the decision to submit it for publication.

References (29)

  • L. Ayoub et al.

    Implementation of a data management program in a pediatric cancer unit in a low income country

    Pediatr Blood Cancer

    (2007)
  • R.S. Arora et al.

    The problem of treatment abandonment in children from developing countries with cancer

    Pediatr Blood Cancer

    (2007)
  • M. Bonilla et al.

    Prevalence and predictors of abandonment of therapy among children with cancer in El Salvador

    Int J Cancer

    (2009)
  • S.C. Howard et al.

    Improving outcomes for children with cancer in low-income countries in Latin America: a report on the recent meetings of the Monza International School of Pediatric Hematology/Oncology (MISHPO) – part I

    Pediatr Blood Cancer

    (2007)
  • Cited by (30)

    • Reduced-intensity therapy for pediatric lymphoblastic leukemia: Impact of residual disease early in remission induction

      2021, Blood
      Citation Excerpt :

      The results of our study also have important implications for treating ALL in low- to middle-income countries (LMICs). Intensification of ALL therapy, which is associated with the high cure rates seen in high-income countries (HICs), has had minimal impact on ALL outcomes in LMICs because of treatment-related mortality (TRM).29-33 Many centers in LMICs with adequate infrastructure typically adopt the treatment protocols used in HICs.

    • Childhood cancer mortality trends in Europe, 1990-2017, with focus on geographic differences

      2020, Cancer Epidemiology
      Citation Excerpt :

      The cause of geographic disparities is multifactorial, encompassing reduced availability of specialized paediatric oncology centres, changes in diagnostic capacity and improved death registration. Moreover, differences in socioeconomic characteristics may have also contributed to the geographic variability in childhood cancer survival [20–23]. In central and eastern countries, including particularly Serbia, Ukraine and the Russian Federation, and some southern European countries, an important share of the population lives in areas with limited availability of specialized healthcare and supportive services.

    • The magnitude and predictors of abandonment of therapy in paediatric acute leukaemia in middle-income countries: A systematic review and meta-analysis

      2013, European Journal of Cancer
      Citation Excerpt :

      Nine studies defined abandonment, ranging from 4 to 12 weeks of missed therapy.17,19–26 Only two studies actively followed up cases of abandonment to minimise misclassification.21,22 For meta-analysis, all studies reporting ARs that included upfront refusal were assembled and narrowed to one study per institution, leaving 40 studies representing 10,494 children in 20 countries (Supplemental Table 2).

    • The evolution of comprehensive cancer care in Western Kenya

      2013, Journal of Cancer Policy
      Citation Excerpt :

      However, as invaluable as these guideline and recommendations are, implementation has lagged, so there is limited documentation of successful models of resource-appropriate cancer care collaboration between high-income countries and low- and middle-income countries (HC–LMC collaborations) in the literature. The program descriptors and documentation of programmatic evolution in pediatric oncology, mainly from the Monzas International School of Pediatric Hematology/Oncology (MISPHO) and the St. Jude International Outreach Program, (IOP), form the bulk of the published literature [14–19]. More recently, descriptors of the process and implementation of creating a breast-specific cancer care delivery system have been published [20].

    View all citing articles on Scopus
    d

    Contributed equally to this manuscript.

    View full text