Epidemiological Features of Mortality from Acute Leukemias in Children of Chelyabinsk Oblast over a 20-Year Period from 1982 to 2002

The dissertation by Shatalova E.Yu. is devoted to the study of epidemiological features of mortality from acute leukemias in children of Chelyabinsk Oblast over a 20-year period (1982–2002). The research encompasses the analysis of the structure and dynamics of mortality, causes of fatal outcomes, survival indicators, and risk factors for an unfavorable outcome of acute leukemia. Special attention is given to the impact of the adoption of programmed treatment and international chemotherapy protocols (BFM group) on treatment outcomes for children in the region.

The work includes a literature review, a description of materials and methods, an epidemiological analysis, an investigation of causes of death, a factor analysis of risks, and the development of a mathematical model for predicting the disease outcome. The results make it possible to identify treatment stages with the highest frequency of unfavorable outcomes, determine criteria for the formation of risk groups, and create a predictive model for the classification of outcomes in patients with acute lymphoblastic leukemia.

Relevance of the Problem

In the structure of mortality from malignant neoplasms among children and adolescents, the largest proportion is occupied by hematoblastoses — 42% [Shamanskaya T.V. et al. 2005; Pui C.N., 1998]. Acute leukemias (AL) occupy a leading position among oncological diseases of childhood, accounting for 30–40% of all malignant diseases: acute lymphoblastic leukemia (ALL) comprises 80%, and acute non-lymphoblastic leukemia (ANLL) — 15–17% [Zhukovskaya E.V., 2005; Aleynikova O.V., 1999; Gao Y.G., et al. 2006; Viscomi S., Pastore G., Mosso M.L. et al. 2003; Carolyn A., Felix A., Beverly J. 1999].

The incidence of acute leukemias among children in the Chelyabinsk Oblast is 3.7 per 100,000 children [Zhukovskaya E.V., 2005], while the national average is 3.2 per 100,000 children.

According to domestic authors [Trapesnikov N.N., Aksel E.M. 2001], the highest mortality rate from acute leukemia (both sexes) was recorded in the 0–4 age group and amounted to 2.2 per 100,000 children. Nevertheless, the data found in the same literary sources regarding the ranges of the standardized mortality rate from leukemia among children of Chelyabinsk Oblast (for the period 1996–2000) — from 0.74 in 1997 to 1.8 in 1999 per 100,000 children — do not provide a complete picture of the epidemiological, demographic, and territorial features of childhood mortality from this pathology within the region, particularly against the background of the adoption of international standards for acute leukemia therapy since the early 1990s at the hematological center of the Chelyabinsk Oblast Children's Clinical Hospital (OHGTS CHO DKHB).

In the work of N.Yu. Drozhdilova (2006) on the epidemiology of hematoblastoses in Chelyabinsk Oblast for the period 1992–2003, it was demonstrated that there was no significant decline in mortality from hematoblastoses for all age groups of the population, despite the full provision of programmed and supportive therapy, which is inconsistent with mortality trends for acute leukemias in the pediatric population [Kovalenko S.G., 2002; Karachunsky A.V. 1998, 1995; Samochatova E.V. 1997].

Chelyabinsk Oblast is a zone of heightened ecological tension, which gives rise to keen interest in the study of mortality from acute leukemias among children residing in conditions of increased technogenic load: in the territories of closed administrative-territorial formations (ZATO) and in the zone of the Urals nuclear waste storage facility (VURS), in cities with varying levels of urbanization and industrial development [Krestinina L.Yu. 2005; Koshurnikova N.A. 2005; Medvedeva I.A. 2004; Classification and Typology of Cities: Electronic Resource, 2003; A.V. Akleyev 2000; Koshenko M.M. 1996].

Over the past 20 years, there has been a significant improvement in survival outcomes for children with AL. In the 1970s–1980s, as a result of the application of standard treatment regimens, the 5-year survival rate in our country for ALL did not exceed 20%, and for ANLL — 5%. However, after the adoption in the 1990s of intensive chemotherapy programs (protocols ALL-BFM-86, 90, 95m; AML-BFM-87, 93, 98), these indicators increased substantially: the 5-year event-free survival for ALL reached 70–75%, and the 3-year event-free survival for ANLL — 40–50% [Popa A.V., 2006; Zhukovskaya E.V., 2005; Migali A.V., 2003; Kovalenko S.G., 2002; Karachunsky A.I., 1999; Rumyantsev S.A., 1997]. These achievements were attained largely through risk-adapted (taking into account risk factors) intensification of chemotherapy, implementation of programmed treatment, utilization of the latest developments in supportive therapy, and the introduction of bone marrow transplantation in children with a poor disease prognosis [Urmaeva M.M., 1998; Rumyantsev S.A., 1997; Rumyantsev A.G. 1994]. Nevertheless, a significant number of patients remain with therapy-resistant variants of ALL; patients who succumbed to complications during the first weeks of induction or even remission [Peshikova M.V., 2004; Rumyantsev A.G. 1994].

The first data on the structure of causes of fatal outcomes in children were obtained from leukemia studies based on Moscow clinics [Urmaeva M.M., 1998; Rumyantsev S.A., 1997].

The study of protocol-based treatment of AL in children of Chelyabinsk Oblast is limited to the investigation of infectious complications [Peshikova M.V., 2004]. Therefore, an analysis of the causes of death in children of our oblast from AL across different treatment periods, taking into account the diverse morphological and immunological properties of the tumor clone and various clinical manifestations, prompts an interpretation of the features of fatal outcomes at the regional level, followed by a practical approach to treatment.

The result of treatment is the achievement of a certain outcome: remission, relapse, or fatal outcome, which forms the basis for calculating various types of survival (overall, relapse-free, event-free) in patients.

Various literary sources [Analysis of Risk Factors: Electronic Resource, 2004; Chessells JM., 1995] indicate ever-new risk factors for developing acute leukemia, but risk factors for an unfavorable outcome, beyond the standard set incorporated in the protocol, could not be found in the available literature. The question of studying risk factors for fatal outcomes in children with acute leukemia remains open to this day.

Over the past decade, a very limited number of studies [Zhukovskaya E.V., 2005; Peshikova M.V., 2004; Kovalenko S.G., 2002] have been registered in Chelyabinsk Oblast, devoted to the study of childhood mortality indicators from acute leukemias, causes of death, especially over long time periods (10–20–30 years) [Drozhdilova N.Yu., 2006], including against the background of changed approaches to treatment (adoption of modern protocols and strengthening of supportive therapy). Therefore, the calculation of mortality and survival indicators in patients with AL over a long time interval (especially after changes in treatment approaches), risk assessment, and the creation of a predictive model for the outcome of ALL in children of Chelyabinsk Oblast are relevant, representing a reliable tool for improving specialized care for children, which can serve as a basis for adequate planning of the work of the hemato-oncological service, including at the regional level.

Aim

To study the regional features of childhood mortality from acute leukemias in the territory of Chelyabinsk Oblast for the period 1982–2002 and to conduct a comparative analysis by decades.

Research Objectives

1. To study the epidemiological features of certain mortality indicators from acute leukemia among children of Chelyabinsk Oblast for the period 1982–2002.

2. To identify the features of causes of death in children with acute leukemia in the second decade at the regional level.

3. To analyze the dynamics of survival in children with acute leukemia (before and after the adoption of BFM group protocols).

4. To identify factors significantly influencing the likelihood of a fatal outcome in children with acute leukemia through the calculation of risks (relative risk, odds ratio, percentage of attributable risk).

5. To determine complex risk factors with the greatest contribution to disease outcomes (died; survived) in children with acute lymphoblastic leukemia using factor analysis.

6. To create a model for calculating the probable outcome of the disease (died; survived) in patients with acute lymphoblastic leukemia.

Scientific Novelty

For the first time, the epidemiological features of certain mortality indicators in acute leukemia among children aged 0–14 in Chelyabinsk Oblast for the period 1982–2002 and by decades were studied, taking into account age, sex, place of residence, climatic-geographic zone, and category of urban settlements. A statistically significant reduction in the frequency of fatal outcomes was revealed, and a significant increase in survival of patients with acute lymphoblastic leukemia was obtained following the adoption of programmed polychemotherapy (PCT).

Risk factors for an unfavorable outcome of acute leukemia of medium and maximum prognostic significance were established.

A statistically significant decrease in the frequency of fatal outcomes from leukemia progression in the second decade was observed.

As a result of the conducted analysis of causes of death, treatment stages of acute leukemia with the highest frequency of unfavorable outcomes depending on the presence or absence of aplasia of hematopoiesis were identified, on the basis of which the formation of groups at increased risk of fatal outcomes in patients was proposed.

A mathematical model for predicting the development of a fatal outcome in acute lymphoblastic leukemia in children of Chelyabinsk Oblast was developed.

Practical Significance

Data obtained from the calculation of the percentage of attributable risk made it possible to identify criteria for the formation of risk groups for the occurrence of an unfavorable outcome in AL in children, which will facilitate an individualized approach to therapy.

The identification of treatment stages of AL with the highest frequency of fatal outcomes depending on the presence (relapse period, cytoreductive phase, remission status) or absence of aplasia of hematopoiesis (first three days from the moment of diagnosis, relapse period, or remission status) will make it possible to form patient groups at increased risk of fatal outcomes in order to optimize organizational and therapeutic approaches aimed at improving patient survival and reducing mortality indicators at the regional and federal levels.

The developed mathematical model for predicting the outcome of ALL in children will help identify at an early stage patients with an unfavorable disease prognosis, which will require strict adherence to polychemotherapy regimens; implementation of supportive therapy using reserve antibacterial and antifungal agents, timely initiation of hemodynamic correction, and replacement therapy in sufficient volume; in the case of therapy refractoriness — an early decision on alternative treatment methods.

Main Propositions for Defense

1. In the second decade, as a result of the transition to programmed treatment, a statistically significant decrease in the frequency of fatal outcomes (p=0.001) in children with AL was revealed; a significant difference by sex with a predominance of boys was obtained (p=0.001). The peak of fatal outcomes in the first decade falls on the 5–9 age category; in the second decade — 10–14 years, which is explained by the increase in life expectancy after the transition to treatment according to protocols.

2. The adoption of programmed therapy in the second decade made it possible to significantly (p=0.015) reduce the frequency of fatal outcomes from AL due to disease progression. Treatment stages with the highest share of unfavorable outcomes were identified: depending on the presence (relapse period, cytoreductive phase, or remission status) or absence of aplasia of hematopoiesis (first three days from the moment of diagnosis, relapse period, or remission status), which made it possible to form patient groups at increased risk of unfavorable outcomes.

3. The adoption of risk-adapted polychemotherapy led to a statistically significant (Log-Rank test: p=0.001) increase in all types of survival for children with acute lymphoblastic leukemia: overall 5-year survival in the second decade reached 72.60±3.57%; event-free 5-year — 60.30±4.00% (p=0.001). The overall 5-year (79.45±3.42%) and event-free 5-year (65.53±4.10%) survival indicators of patients with acute lymphoblastic leukemia who received only protocol-based treatment exceeded the corresponding survival indicators for the entire study group of children with acute lymphoblastic leukemia in the second decade and were significantly (p=0.001) different from the survival of patients receiving non-protocol-based treatment.

4. On the basis of the analysis of the percentage of attributable risk, criteria for the formation of groups of maximum (%AR=51–100%) and medium (%AR=26–50%) prognostic significance of the risk of an unfavorable outcome in acute leukemia were identified. Factors attributed to the medium prognostic significance risk group for fatal outcomes in children included: initial white blood cell count in peripheral blood >10×10⁹ (%AR=31.83%); height at birth less than 48 cm or more than 53 cm (%AR=35.29%); birth weight less than 2500 g or more than 4000 g (%AR=39.74%). The group of maximum prognostic significance risk of an unfavorable outcome in acute leukemia in children included: premature birth (%AR=51.37%); isolated bone marrow involvement at any relapse (%AR=56.50%); relapses in medical history (%AR=62.94%).

5. Among complex risk factors contributing the most to the outcome of ALL (survived; died), tumor clone volume and oncological burden across generations were identified, with their share in the unfavorable outcome (12.00% and 19.00%, respectively) being predominant.

6. The mathematical model for predicting the outcome of acute lymphoblastic leukemia in children aged 0–14, developed on the basis of discriminant analysis, allows correct classification of the disease outcome (survived; died) in patients of Chelyabinsk Oblast with a probability of 82.6%.

Implementation of Results

The results of the dissertation research have been implemented in the practice of the hemato-oncological center of the State Scientific-Practical Health Institution Chelyabinsk Oblast Children's Clinical Hospital, and are used in the educational process at the Department of Pediatric Diseases No. 2 of the State Educational Institution of Higher Professional Education "Chelyabinsk State Medical Academy" of the Federal Agency for Health and Social Development.

Testing of the Work and Publications

The materials of the dissertation were presented at the III Summary Scientific-Practical Conference of Young Scientists of Chelyabinsk State Medical Academy (Chelyabinsk, 2005); at the III All-Russian Forum "Health of the Nation — the Foundation of Russia's Prosperity" (Moscow, 2007).

Seven printed works have been published on the topic of the dissertation.

Scope and Structure of the Work

The dissertation consists of an introduction, five chapters, a conclusion, findings, practical recommendations, and a bibliography containing 217 sources, of which 121 are domestic and 96 are foreign. The work comprises 205 pages of typescript, is illustrated with 43 tables and 16 figures, and includes 2 appendices.