ACUTE LEUKEMIA OF CHILDHOOD-A SINGLE INSTITUTION ' S EXPERIENCE

The aim of this study was to investigate distribution of immunophenotypic and cytogenetic features of childhood acute leukemia (AL) in the cohort of 239 newly diagnosed patiens registered at the leading pediatric oncohematology center in the country during a six-year period (1996-2002). With approximately 60-70% of all childhood AL cases in Serbia and Montenegro being diagnosed and treated in this institution the used data represent a valid research sample to draw conclusions for entire country. On the basis of five phenotypic markers, the distribution of immunological subtypes was as follows: 169 (70.7%) expressed B-cell marker CD19 (137 were CD10 positive and 32 CD10 negative), 37 (15.5%) belonged to T-lineage acute lymphoblastic leukemia (T-ALL) (cyCD3 positive), and 33 (13.8%) were acute myeloblastic leukemia (AML) (CD13 positive and/or CD33 positive in the absence of lymphoid-associated antigens). The ratio of males and females was 1.5:1. Most of the cases were between the ages of 2 and 4, and were predominantly B-lineage acute lymphoblastic leukemia (B-ALL) cases. Another peak of age distribution was observed at the age of 7. The frequency of T-ALL (18% of ALL) was similar to that reported for Mediterranean countries: France (19.4%), Greece (28.1%), Southern Italy (28.3%), and Bulgaria (28.0%). Cytogenetic analyses were performed in 193 patients: 164 ALL and 29 AML. Normal karyotype was found in 57% of ALL and in 55% of AML patients, while cytogenetic abnormalities including structural, numerical, and complex chromosomal rearrangements were found in 43% of ALL and in 45% of AML patients. Our results represent a contribution to epidemiological aspects of childhood leukemia studies. UDC 616.155.392 053.2(497.11)


INTRODUCTION
The acute leukemias, the most frequent type of childhood cancer in developed countries, are a heterogenous group of diseases with regard to type, treatment response, prognosis, and (probably) etiology as well (Parkin et al. 1998;Pui, 1999).This variation, combined with the relative rarity of the diseases, compromises epidemiological studies aimed at identifying causal associations.Etiological clues could (however) be derived from comparing national incidence rates.Incidence rates of AL, its age dependence, and geographic variations within the different subgroups might be linked with socioeconomic status and concomitant exposure to potential leukemogenic agents (Greaves et al. 1993;Ramotand Magrath, 1982).As a contribution to such epidemiological studies, we here present systemized data on the distribution of immunophenotypic and cytogenetic features of AL subgroups in Serbia and Montenegro.The research is based on cases registered at the Department of Oncohematology's Pediatric Clinic of the Mother and Child Health Institute of Serbia in Belgrade.The department is the leading pediatric oncohematology center in the country.With approximately 70% of all childhood AL cases in Serbia and Montenegro being diagnosed and treated in this institution, the used data represent a valid research sample from which to draw conclusions for the entire country.

Patients and immunophenotype
From October 1996 to May 2002, AL was diagnosed in 260 patients admitted to the Mother and Child Health Institute of Serbia.Immunophenotyping was performed on a series of 250 previously untreated patients, and successful results were obtained in 239 of them.The expression of CD19, CD10, cytoplasmic CD3 (cyCD3), CD13, and CD33 antigens was assessed by the alkaline-phosphatase/antialkaline-phosphatase (APAAP) method (Erber et al. 1986;Erberand Mason,1985).All monoclonal antibodies as well as the APAAP kit were purchased from Dako (Dakopatts, Glostrup, Denmark).

Morphological, cytochemical, and cytogenetic analyses
All cases met the conventional morphological and cytochemical criteria (Sudan BB and ANAE) for the diagnosis and were further classified according to the FAB system (Van den Does-van den Berg et al. 1992;Benett et al. 1976).Cytogenetic analysis was performed on bone marrow cells after direct preparation and/or unstimulated 24-48 h short-term culture.At least 20 metaphases were analyzed for each case, using conventional cytogenetics with GTG banding.

Age-related distribution of AL patients
The age distribution is shown in Fig. 4. The median age for ALL was 5 (range: 4 months to 17 years) and for AML 7 (range: one day to 16 years).The overall peak of incidence was between the ages of 2 and 4.This peak was formed predominantly by B-ALL CD10+, which represented 87% of all cases in this age group, where another peak at the age of 7 was observed.After the age of 7, the proportions of B and T subtypes were similar.The peak incidence of T-ALL was between the ages of 8 and 13.We noticed a bimodal distribution of AML with major peaks at the ages of 1 and 8.

DISCUSSION
In the present study we have analyzed the immunophenotype of 239 patients and cytogenetic features of 228 patients with childhood AL referred to a single institution in Serbia and Montenegro.Our series allowed us to draw conclusions about the distribution of childhood AL in the entire country, as 60-70% of childhood acute leukemia cases from all regions were treated in this clinic.
The majority of cases (71%) were precursor B-ALL (either CD10+ or CD10-).The age distribution showed a clear peak between the ages of 2 and 4 with a predominance of precursor B-ALL CD10+.The results on age and sex distribution and phenotypic pattern of our patients were in agreement with those reported by other authors for Caucasian children (Parkin et al. 1998;Greaves et al. 1993).However, some differences were observed.First, the precursor B-ALL CD10+ displays a peak of age distribution at the age of 7, which does not fully correspond to the data from either developed or undeveloped countries.It is well documented in previous studies that earlier in the 20th century in developed countries there was an earlier appearance of the 2-5 year peak of ALL, as there was some positive association between socio-economic status and ALL incidence (Erberand Mason, 1985).The typical 2-5 year peak is evident in Serbia and Montenegro, but there is a higher frequency of patients of age 5-10 and ≥10 than was observed in developed countries (Malta et al. 1997).Secondly, our results indicate that the proportion of T-ALL (18%) is between the groups of countries with low and moderate proportion of T-ALL (Papamichail et al. 1985;Taskov et al. 1995;Russo1985).We could not ascertain it fully, but our assumption is that our findings could be explained by influence of environmental factors: socio-economic development and concomitant exposure to potential leukemogenic agents (Greaves et al. 1993;Ramot andMagrath, 1982).The socioeconomic circumstances in the region (West Balkans) are important for the proper interpretation of this study's results.This region has always been relatively underdeveloped by European standards, but the socio- PDF created with FinePrint pdfFactory Pro trial version www.pdffactory.comeconomic situation dramatically deteriorated from the beginning of the 1990s and the outbreak of the wars for the Yugoslav succession.The already grave humanitarian situation was additionaly complicated by significant demographic changes: Serbia was flooded by almost one million refugees from Serbian-populated regions of Croatia and Bosnia and Herzegovina and internally displaced persons from Kosovo and Metohija.An additional factor that remains to be carefully monitored and researched is the potential influence of toxic substances the population was exposed to during the war devastations (e.g., depleted uranium and spills from bombed-out chemical industrial facilities and oil refineries).We plan to collect detailed epidemiological data and characterize time trends in the age-and sex-specific incidence of childhood AL in the future.Such epidemiological studies have already been done in developed countries (Hjalgrim et al. 2003).
Cytogenetic abnormalities, both lymphoblastic and myeloid, are common in patients with AL, and some of these have prognostic significance.The most frequent translocations in ALL are: t(1;19), t(4;11), others involving the 11q23 breakpoint, and t(9;22) in B-ALL; t(8;14), t(2;8), and t(8;22) in ALL with L3 morphology; and t(11;14) and (1;14) in T-ALL.In the group of 35/164 ALL with structural (18 cases) and complex (17 cases) chromosomal rearrangments, common translocations were found in nine cases: 7 B-ALL [t(9:22) in five cases; t(8;14) in one case; t(4;11) in one case] and 2 T-ALL [t(1;14) in one case and t(11;14) in one case].In the remaining 26 ALL patients, non-specific structural aberrations (translocations, inversions, deletions, and duplications) involving chromosomes 1, 9, 3, 19, and 18, were found.Some of the common translocations described above are recognized as predictive of outcome: t(9;22) is present in approximately 3%-4% of pediatric ALL patients having an unfavorable prognosis, which is in line with our results describing five cases of t(9;22) in the group of 164 ALL patients (3%) (Hann et al. 2001;Arico et al. 2000;Schrappe et al. 1998;Ribeiro et al. 1997);t(4;11) and others involving 11q23 occur in approximately 4% of cases, which usually show a poor response to initial therapy (Rubnitzand Pui, 1987;Rubnitzand Look, 1998).In the analyzed group of 164 ALL, these translocations were found in 2/164 ALL cases, both having B-ALL phenotype (1%).Hyperdiploidy (>50 chromosomes per cell or DNA index>1.16)occurs in 20% to 25% of cases of B-precursor ALL, but very rarely in cases of T-cell ALL (Puiand Evans, 1998) and is associated with a favorable prognosis.In our group of 164 ALL patients, hyperdiploidy (47-57 chromosomes) was found in 26/164 cases (16%) both B cell lineage ALL.In eight cases of B-ALL near triploidy (58-80 chromosomes) was found and in 2 other cases of B-ALL near tetraploidy (92± chromosomes) was observed.
We started flow cytometric immunophenotyping of childhood AL in 2002.Today, we have a well documented diagnostics for the entire nation (until November 2002 about 10-20% of the population was not covered).We expect to have complete national data for imunophenotype and cytogenetic features of childhood AL in the near future.

Fig 1 .
Fig 1. Distribution of immunophenotypes of childhood AL in 239 patiens in Serbia and Montenegro

Fig. 4 .
Fig. 4. Age-related distribution of childhood AL in 239 patients in Serbia and Montenegro

Table 1 .
Age and sex distribution of immunophenotypes of childhood AL in 239 patients in Serbia and Montenegro