Abstract
Background
Pediatric cancer survival has improved substantially worldwide; however, evolving incidence patterns and treatment-related toxicity continue to challenge health systems, particularly in transitioning settings. This study evaluated population-level trends and clinical toxicity patterns of osteosarcoma, neuroblastoma, and acute lymphoblastic leukemia (ALL) in Kazakhstan.
Methods
A retrospective population-based epidemiological analysis was combined with a hospital-based clinical cohort study. Incidence and mortality data were obtained from the national cancer registry maintained by the Scientific Center for Health Development for 2016–2025. Temporal trends were analyzed using Poisson regression, and annual percent change (APC) was calculated. Forecasts for 2025–2030 were generated using regression-based models. Clinical data on treatment-related toxicity were retrospectively collected from medical records of patients treated at the National Scientific Center of Pediatrics and Pediatric Surgery. Multivariable logistic regression models were used to identify predictors of severe hematologic toxicity and related clinical outcomes.
Results
Between 2016 and 2025, the incidence increased significantly for osteosarcoma (APC 18.2%, 95% CI 8.8–28.4; p = 0.0001) and neuroblastoma (APC 10.3%, 95% CI 5.0–15.9; p = 0.0001), while ALL demonstrated a non-significant upward trend. Mortality remained statistically stable across all malignancies. Forecast models suggest continued growth in incidence through 2030. In clinical cohorts, severe treatment-related toxicity was common. Advanced disease stage was associated with severe toxicity in osteosarcoma, while risk group was the primary determinant of hematologic toxicity in ALL. In neuroblastoma, toxicity was most pronounced among patients with high-risk disease receiving intensive chemotherapy.
Conclusion
Incidence of osteosarcoma and neuroblastoma is increasing in Kazakhstan while mortality remains stable. Despite stable survival, the burden associated with osteosarcoma, neuroblastoma, and acute lymphoblastic leukemia is increasing due to rising case numbers and substantial treatment-related toxicity.
Supplementary Information
The online version contains supplementary material available at https://doi.org/10.1186/s12885-026-16570-6.
Keywords: Pediatric cancer, Osteosarcoma, Neuroblastoma, Acute lymphoblastic leukemia, Incidence trends, Mortality, Treatment-related toxicity, Forecast modeling, Kazakhstan, Health policy
Introduction
Each year, approximately 400,000 children and adolescents (0–19 years) are diagnosed with cancer worldwide. The most common malignancies include leukemias, brain tumors, lymphomas, and embryonal tumors such as neuroblastoma. Survival exceeds 80% in high-income countries but remains below 30% in many low- and middle-income settings, reflecting disparities in access to timely diagnosis and specialized care [1].
In Kazakhstan, pediatric oncology represents a growing public health priority. Nationwide data from 2014 to 2021 identified 4,831 children with hematological malignancies, demonstrating a higher incidence of acute lymphoblastic leukemia (ALL) compared with acute myeloid leukemia and relatively stable all-cause mortality rates (1.1–1.5 per 100,000) [2]. Mortality risk was significantly higher among children aged 0–5 years with acute myeloid leukemia. According to the WHO Cancer Country Profile (2020), approximately 586 new cancer cases occur annually among children aged 0–14 years in Kazakhstan [1, 3]. National cancer control programs, pathology services, and pediatric oncology centers are operational, and childhood cancer survival has reached approximately 75%, with treatment provided free of charge at specialized national centers.
Advances in multimodal therapy, including risk-adapted chemotherapy, radiotherapy, and hematopoietic stem cell transplantation, have substantially improved survival in childhood cancers, particularly ALL [4, 5]. However, evolving incidence patterns and the persistent burden of treatment-related toxicity continue to pose clinical and epidemiologic challenges [6, 7].
Osteosarcoma, neuroblastoma, and ALL are biologically and clinically distinct pediatric malignancies with differing age distributions and therapeutic strategies. Although ALL achieves survival rates exceeding 80% in high-income settings, osteosarcoma and high-risk neuroblastoma require intensive regimens associated with significant toxicity [8, 9]. Rising incidence accompanied by stable mortality may reflect therapeutic progress but may also obscure substantial treatment-related morbidity affecting long-term outcomes.
Therefore, this study aims to (1) evaluate population-level incidence and mortality trends for osteosarcoma, neuroblastoma, and ALL from 2016 to 2025, (2) project patterns through 2030, including sex-specific modeling, and (3) assess treatment-related toxicity and its predictors in disease-specific pediatric cohorts in Almaty city of Kazakhstan.
Methods
Study design and setting
This study comprised a retrospective population-based epidemiological analysis combined with a hospital-based clinical cohort study. Incidence and mortality data were obtained from the National Scientific Center for Health Development (Almaty, Kazakhstan) and covered the period from January 1, 2016, to December 31, 2025. Clinical treatment-related data were collected from medical records of patients treated at the Scientific Center of Pediatrics and Pediatric Surgery. The study included pediatric patients diagnosed with osteosarcoma, neuroblastoma, or ALL.
Data sources
Population-Based Registry Data: Annual incidence and mortality data were extracted from the national cancer registry maintained by the National Scientific Center for Health Development. Extracted variables included: Year of diagnosis; Year of death; Cancer diagnosis (ICD-10 code); Sex; Age at diagnosis; Underlying cause of death. Population denominators were obtained from official national demographic statistics to support temporal modeling.
Clinical cohort data
Clinical data were collected retrospectively from medical records of patients treated at the National Center of Pediatrics. Extracted variables included: Age at diagnosis; Sex; Risk group classification; Treatment phase or block; Baseline laboratory parameters (including initial leukocyte count); Occurrence and grade of hematologic toxicity; Infectious complications; Interruption of maintenance therapy.
Hematologic toxicity was graded according to the Common Terminology Criteria for Adverse Events (CTCAE). Severe toxicity was defined as grade ≥ 3.
Case definition and classification
Cancer cases were identified using the International Classification of Diseases, 10th Revision (ICD-10).
Osteosarcoma was defined using codes C40 (malignant neoplasm of bone and articular cartilage of limbs) and C41 (malignant neoplasm of bone and articular cartilage of other and unspecified sites). Neuroblastoma cases were identified using codes C74 (adrenal gland), C48 (retroperitoneum and peritoneum), C38 (mediastinum), and C47 (peripheral nerves and autonomic nervous system). ALL was defined using ICD-10 code C91.0.
Only newly diagnosed primary malignant cases were included in incidence analyses. Mortality analyses were based on cancer-specific underlying cause of death recorded in the registry.
Epidemiological outcomes included
Annual cancer incidence and mortality. Clinical Outcomes included: Grade ≥ 3 hematologic toxicity; Infectious complications; Maintenance therapy interruption.
Statistical analysis
Temporal trends in incidence and mortality were analyzed using Poisson regression models. Annual percentage change (APC) was calculated from regression coefficients using the formula:
 |
Negative binomial regression models were applied when overdispersion was detected.
Sex-specific trend modification was evaluated by including interaction terms (Year × Sex).
Model linearity was assessed using restricted cubic spline functions, and likelihood ratio tests were performed to compare linear and non-linear models.
Forecasts for 2025–2030 were generated using fitted regression models under the assumption that observed temporal trends would continue. Predicted annual case numbers and 95% confidence intervals were calculated. Increasing uncertainty over time was reflected by widening confidence intervals.
Multivariable logistic regression models were used to identify predictors of: Hematologic toxicity; Infectious complications; Maintenance therapy interruption. Covariates included age, sex, risk group, and baseline leukocyte count. Adjusted odds ratios (AORs) with 95% confidence intervals were reported. Crude models were calculated separately and presented in Supplementary Tables.
Model diagnostics
Overdispersion was assessed using dispersion parameters. Multicollinearity was evaluated using variance inflation factors. Goodness-of-fit was assessed using deviance statistics. Sensitivity analyzes were conducted to confirm the robustness of linear trend assumptions.
Results
Population-level incidence and mortality trends (2016–2025)
Between 2016 and 2025, significant increases in incidence were observed for osteosarcoma and neuroblastoma, whereas ALL demonstrated a non-significant upward trend.
Osteosarcoma incidence increased by an APC of 18.2% (95% CI 8.8–28.4; p = 0.0001). Neuroblastoma incidence increased by 10.3% annually (95% CI 5.0–15.9; p = 0.0001). In contrast, ALL showed a non-significant increase (APC 28.4%, 95% CI − 2.8 to 69.8; p = 0.0787), Table 1.
Table 1.
Temporal trends, sex differences, and forecasted incidence of pediatric osteosarcoma, neuroblastoma, and acute lymphoblastic leukemia
| Outcome | APC % (95% CI) | p-value | Sex Interaction IRR (95% CI) | p-int | Non-linearity p | Forecasted Annual Incidence (Total, 95% CI) | Male Trend β (SE), p | Level Difference F–M β (SE), p | Trend Difference F–M β (SE), p | R² | Adj R² |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Incidence | |||||||||||
| Osteosarcoma | 18.2 (8.8–28.4) | 0.0001 | 1.07 (0.91–1.26) | 0.457 | 0.78 | 177.5 (111.5–243.5) | 4.48 (2.27), p = 0.066 | 12,268 (6,498), p = 0.077 | −6.08 (3.22), p = 0.077 | 0.30 | 0.17 |
| Neuroblastoma | 10.3 (5.0–15.9) | 0.0001 | — | — | — | 174.4 (94.8–254.0) | 3.38 (2.80), p = 0.245 | 11,638 (7,999), p = 0.165 | −5.77 (3.96), p = 0.164 | 0.25 | 0.10 |
| ALL | 28.4 (− 2.8–69.8) | 0.0787 | — | — | — | — | −3.91 (4.10), p = 0.354 | −6,762 (11,702), p = 0.571 | 3.30 (5.79), p = 0.576 | 0.64 | 0.58 |
| Mortality | |||||||||||
| Osteosarcoma | −3.79 (− 21.74–18.28) | 0.714 | — | — | — | — | — | — | — | — | — |
| Neuroblastoma | 6.84 (− 3.29–18.02) | 0.193 | — | — | — | — | — | — | — | — | — |
| ALL | −1.21 (− 9.33–7.65) | 0.782 | — | — | — | — | — | — | — | — | — |
| Forecasted incidence (2025–2030) | |||||||||||
| All malignancies (combined) | — | — | — | — | — | 588.7 (502.0–675.4) | — | — | — | — | — |
APC annual percent change, IRR— incidence rate ratio, CI confidence interval, SE standard error, F–M— female minus male, ALL acute lymphoblastic leukemia
Despite the rising incidence, mortality remained statistically stable across all malignancies. APC estimates were − 3.79% (p = 0.714) for osteosarcoma, 6.84% (p = 0.193) for neuroblastoma, and − 1.21% (p = 0.782) for ALL (Table 1; Figs. 1 and 2).
Fig. 1.
Projected Cancer Incidence Through 2030
Fig. 2.
Projected Cancer Mortality Through 2030
Projected incidence and mortality (2025–2030)
Projections suggest continued increases in osteosarcoma and neuroblastoma incidence through 2030, whereas mortality is expected to remain statistically unchanged. By 2030, osteosarcoma incidence is projected to reach approximately 17–18 annual cases in both sexes. Neuroblastoma incidence is projected to exceed 25 annual cases among females. For ALL, projections indicate relative stabilization with wide confidence intervals reflecting uncertainty. Mortality projections across all malignancies demonstrate modest fluctuations without statistically significant directional change. Confidence intervals widen over the forecast horizon, indicating increasing uncertainty inherent to time-series extrapolation (Supplementary Table 1; Figs. 1 and 2).
Sex-specific forecast models showed higher projected incidence among males for osteosarcoma (96.0 vs. 81.5 cases annually) and neuroblastoma (97.1 vs. 77.3 cases), although sex-based level and trend differences were not statistically significant (all p > 0.05). No significant sex differences were observed for ALL. Model fit was modest for osteosarcoma (R²=0.30) and neuroblastoma (R²=0.25), and stronger for ALL (R²=0.64). Combined projected annual incidence reached 338.3 cases in males and 250.4 cases in females (Table 1).
Clinical cohort analyses
To complement population-level findings, we evaluated treatment-related toxicity and clinical predictors within disease-specific pediatric cohorts.
Osteosarcoma cohort
A total of 38 patients with osteosarcoma were included, with a median age of 13.5 years (IQR 11.6–15.0). Males predominated (63.2%). Among patients with available surgical data (n = 21), complete tumor resection (R0) was achieved in the majority (95.2%). Stage information was available for 23 patients, and more than half presented with advanced disease (IVA–IVB, 56.5%). According to TNM classification (n = 25), most tumors were classified as T2 (76.0%), while nodal involvement was rare (N1, 4.0%). Distant metastases were documented in 24.0% of patients. A total of 278 chemotherapy blocks were administered, predominantly AP-containing regimens (57.6%), followed by MM-containing regimens (35.3%), Table 2.
Table 2.
Clinical and demographic characteristics of patients with osteosarcoma (n = 38)
| Variable | Value |
|---|---|
| Age, median (IQR) | 13.5 (11.6–15.0) |
| Sex, n (%) | |
| Male | 24 (63.2) |
| Female | 14 (36.8) |
| Resection status (n = 21), n (%) | |
| R0 | 20 (95.2) |
| Rx | 1 (4.8) |
| Stage (n = 23), n (%) | |
| IA | 4 (17.4) |
| IB | 1 (4.3) |
| IIB | 5 (21.7) |
| IVA | 12 (52.2) |
| IVB | 1 (4.3) |
| Stage (grouped), n (%) | |
| Early (IA–IIB) | 10 (43.5) |
| Late (IVA–IVB) | 13 (56.5) |
| Tumor stage T (n = 25), n (%) | |
| T1 | 5 (20.0) |
| T2 | 19 (76.0) |
| T3 | 1 (4.0) |
| N stage (n = 25), n (%) | |
| N0 | 10 (40.0) |
| N1 | 1 (4.0) |
| Nx | 14 (56.0) |
| M stage (n = 25), n (%) | |
| M0 | 11 (44.0) |
| M1 | 6 (24.0) |
| M1a | 5 (20.0) |
| M1b | 2 (8.0) |
| Mx | 1 (4.0) |
| Chemotherapy blocks (n = 278), n (%) | |
| AP-containing | 160 (57.6) |
| MM-containing | 98 (35.3) |
| Other | 20 (7.2) |
Treatment-related toxicities were generally mild for most organ systems. Nephrotoxicity, neurotoxicity, cardiotoxicity, and respiratory toxicity were uncommon and predominantly grade 0–1. In contrast, clinically relevant toxicities were observed for hepatotoxicity and hematologic toxicity. Severe hepatotoxicity (grade ≥ 3) occurred in 10 patients (26.3%), while severe hematologic toxicity was observed in 10 patients (26.3%), including grade 4 events in 18.4%. Dermatologic toxicity of grade 3 was also relatively frequent (23.7%).
Analysis of factors associated with severe toxicity (grade 3–4) showed that advanced disease stage was significantly associated with a higher risk of severe toxicity (p = 0.026). In contrast, age, sex, and resection status were not significantly associated with toxicity severity. Notably, patients with late-stage disease demonstrated a substantially higher proportion of severe toxicity compared with those with early-stage disease, Table 3.
Table 3.
Treatment toxicity and factors associated with severe toxicity in patients with osteosarcoma
| Variable | Grade 0 | Grade 1 | Grade 2 | Grade 3 | Grade 4 | |
|---|---|---|---|---|---|---|
| Nephrotoxicity | 34 (89.5) | 3 (7.9) | 1 (2.6) | 0 | 0 | |
| Hepatotoxicity | 13 (34.2) | 6 (15.8) | 9 (23.7) | 9 (23.7) | 1 (2.6) | |
| Dyspeptic syndrome | 19 (50.0) | 11 (28.9) | 7 (18.4) | 1 (2.6) | 0 | |
| Neurotoxicity | 37 (97.4) | 0 | 0 | 0 | 1 (2.6) | |
| Cardiotoxicity | 38 (100) | 0 | 0 | 0 | 0 | |
| Dermatologic toxicity | 29 (76.3) | 0 | 0 | 9 (23.7) | 0 | |
| Coagulologic toxicity | 37 (97.4) | 0 | 0 | 1 (2.6) | 0 | |
| Respiratory toxicity | 38 (100) | 0 | 0 | 0 | 0 | |
| Hematologic toxicity | 15 (39.5) | 9 (23.7) | 4 (10.5) | 3 (7.9) | 7 (18.4) | |
| Metabolic disorders | 22 (57.9) | 14 (36.8) | 1 (2.6) | 1 (2.6) | 0 | |
| Factors associated with severe toxicity | ||||||
|---|---|---|---|---|---|---|
| Variable | Toxicity 0–2 (n = 16) | Toxicity 3–4 (n = 22) | p | |||
| Age, median (IQR) | 13.31 (9.25–14.99) | 13.55 (12.01–14.60) | 0.421 | |||
| Male | 10 (41.7) | 14 (58.3) | 0.943 | |||
| Female | 6 (42.9) | 8 (57.1) | ||||
| Early stage | 6 (60.0) | 4 (40.0) | 0.026 | |||
| Late stage | 2 (15.4) | 11 (84.6) | ||||
| R0 | 8 (40.0) | 12 (60.0) | 0.421 | |||
| Rx | 0 | 1 (100) | ||||
Data are presented as median (IQR) or n (%). P-values were calculated using Mann–Whitney U test and χ²/Fisher’s exact test
Acute lymphoblastic leukemia cohort
A total of 215 patients with acute lymphoblastic leukemia were included. The median age was 5.0 years (IQR 3–12), and males predominated (59.5%). Most patients were classified as standard risk (67.9%), and B-lineage ALL was the most common immunophenotype (82.8%), while T-lineage ALL accounted for 7.0%. Additional baseline clinical characteristics are presented in Table 4.
Table 4.
Baseline clinical and demographic characteristics of patients with acute lymphoblastic leukemia (n = 215)
| Variable | Value, median (IQR) or n (%) | |
|---|---|---|
| Age, years | 5.0 (3–12) | |
| Sex | Male | 128 (59.5) |
| Female | 87 (40.5) | |
| Risk group | High/intermediate | 69 (32.1) |
| Standard | 146 (67.9) | |
| FAB classification | L1 | 44 (20.5) |
| L2 | 70 (32.6) | |
| Missing | 101 (47.0) | |
| Immunophenotype | B-lineage ALL | 178 (82.8) |
| T-lineage ALL | 15 (7.0) | |
| MPAL / other | 22 (10.2) | |
| CNS status | CNS1 | 170 (79.1) |
| CNS2 | 6 (2.8) | |
| CNS3 | 3 (1.4) | |
| Missing | 36 (16.7) | |
| Initial leukocyte count (×10⁹/L) | 5.41 (3.12–9.24) | |
| Treatment interruption | Yes | 74 (34.4) |
| No | 65 (30.2) | |
| Anemic syndrome | Yes | 167 (77.7) |
| No | 48 (22.3) | |
| Hepatomegaly | ≤ 5 cm | 16 (7.4) |
| > 5 cm | 87 (40.5) | |
| No | 112 (52.1) | |
| Splenomegaly | ≤ 5 cm | 57 (26.5) |
| > 5 cm | 27 (12.6) | |
| No | 131 (60.9) | |
| Lymphadenopathy | None | 148 (68.8) |
| 1–2 groups | 30 (14.0) | |
| Generalized | 37 (17.2) | |
| HSCT | Yes | 16 (7.4) |
| No | 194 (90.2) | |
Treatment-related toxicities were frequent, with hematologic toxicity being the most severe. Grade 3–4 hematologic toxicity occurred in 72.0% of patients. The full distribution of treatment-related toxicities according to CTCAE grade is shown in Supplementary Table S2.
In multivariable ordinal logistic regression analysis, risk group was the only factor significantly associated with severe hematologic toxicity. Patients with high or intermediate risk had a higher likelihood of developing severe toxicity compared with those in the standard-risk group (AOR 2.54, 95% CI 1.47–4.37; p = 0.001) (Table 5). Results of the crude regression models are provided in Supplementary Table S3.
Table 5.
Predictors and clinical consequences of severe hematologic toxicity
| Section A. Predictors of severe hematologic toxicity (ordinal logistic regression) | ||||
|---|---|---|---|---|
| Variable | AOR (95% CI) | p-value | OR (95% CI) | p-value |
| Age | 0.98 (0.93–1.03) | 0.355 | 0.99 (0.94–1.04) | 0.690 |
| Male sex | 1.11 (0.66–1.85) | 0.693 | 1.07 (0.65–1.75) | 0.795 |
| Initial leukocyte count | 0.99 (0.95–1.03) | 0.511 | 0.98 (0.94–1.02) | 0.324 |
| High/intermediate risk | 2.13 (1.20–3.72) | 0.009 | 2.54 (1.47–4.37) | 0.001 |
| Section B. Clinical outcomes associated with severe hematologic toxicity | ||||
|---|---|---|---|---|
| Variable | Toxicity 0–2 (n = 45) | Toxicity 3–4 (n = 170) | p-value | |
| Weight loss | 0 | 29 (64.4) | 100 (58.8) | 0.157 |
| 1 | 10 (22.2) | 60 (35.3) | ||
| 2 | 4 (8.9) | 8 (4.7) | ||
| 3 | 2 (4.4) | 2 (1.2) | ||
| Karnofsky/Lansky index | 0 | 38 (84.4) | 74 (43.5) | < 0.001 |
| 1 | 5 (11.1) | 71 (41.8) | ||
| 2 | 2 (4.4) | 22 (12.9) | ||
| 3 | 0 | 3 (1.8) | ||
| Infections | 0 | 28 (62.2) | 18 (10.6) | < 0.001 |
| 1 | 10 (22.2) | 12 (7.1) | ||
| 2 | 2 (4.4) | 13 (7.6) | ||
| 3 | 5 (11.1) | 113 (66.5) | ||
| 4 | 0 | 14 (8.2) | ||
| Anemic syndrome | Yes | 38 (84.4) | 129 (75.9) | 0.220 |
| No | 7 (15.6) | 41 (24.1) | ||
| HSCT | No | 44 (97.8) | 150 (90.9) | 0.057 |
| Yes | 1 (2.2) | 15 (9.1) | ||
| Treatment interruption | Yes | 33 (82.5) | 41 (41.4) | < 0.001 |
| No | 7 (17.5) | 58 (58.6) | ||
| Hepatomegaly | No | 43 (95.6) | 156 (91.8) | 0.389 |
| Yes | 2 (4.4) | 14 (8.2) | ||
| Splenomegaly | No | 40 (88.9) | 148 (87.1) | 0.742 |
| Yes | 5 (11.1) | 22 (12.9) | ||
| Lymphadenopathy | No | 20 (44.4) | 128 (75.3) | < 0.001 |
| Yes | 25 (55.6) | 42 (24.7) | ||
| Hemorrhagic syndrome | No | 16 (35.6) | 126 (74.1) | < 0.001 |
| Yes | 29 (64.4) | 44 (25.9) | ||
| Extramedullary foci | No | 40 (88.9) | 166 (97.6) | 0.009 |
| Yes | 5 (11.1) | 4 (2.4) | ||
| Risk group | High/intermediate | 39 (86.7) | 107 (62.9) | 0.002 |
| Standard | 6 (13.3) | 63 (37.1) | ||
Severe hematologic toxicity was also associated with adverse clinical outcomes, including higher rates of infectious complications, worse Karnofsky/Lansky performance status, more frequent treatment interruptions, and a higher prevalence of hemorrhagic syndrome (Table 5).
Neuroblastoma cohort
A total of 61 patients with neuroblastoma were included in the study. The median age at diagnosis was 2.18 years (IQR 1.18–3.84), and males slightly predominated (54.1%). Most patients were classified as high risk (78.6%) and presented with advanced disease, with stage IV observed in 82.7% of cases. MYCN amplification and 1p deletion were detected in 18.2% and 9.1% of patients, respectively. Bone marrow and bone metastases were the most frequent metastatic sites. Baseline clinical and demographic characteristics are summarized in Table 6, while detailed clinical and pathological features are provided in Supplementary Table S4.
Table 6.
Baseline clinical characteristics and infection-related toxicity in patients with neuroblastoma (n = 61)
| Variable | Value, median (IQR) or n (%) | |
|---|---|---|
| Laboratory characteristics | Ferritin | 349.05 (277.50–440.25) |
| LDH | 283.00 (249.40–478.50) | |
| Demographic characteristics | Demographic characteristics | |
| Age, years | 2.18 (1.18–3.84) | |
| Male sex | 33 (54.1) | |
| Female sex | 28 (45.9) | |
| Disease characteristics | High-risk group | 44 (78.6) |
| Intermediate risk | 11 (19.6) | |
| Low risk | 1 (1.8) | |
| Stage II | 3 (5.8) | |
| Stage III | 6 (11.5) | |
| Stage IV | 43 (82.7) | |
| MYCN amplification | 10 (18.2) | |
| 1p deletion | 5 (9.1) | |
| Metastatic disease | Bone marrow metastases | 32 (52.5) |
| Bone metastases | 29 (47.5) | |
| Liver metastases | 14 (23.0) | |
| Lung/pleura metastases | 7 (11.5) | |
| CNS metastases | 7 (11.5) | |
| Treatment characteristics | Radiotherapy | 7 (13.7) |
| HSCT performed | — | |
| Infectious complications by toxicity severity | Infection present (toxicity grade 0–2) | 23 (74.2) |
| Infection present (toxicity grade 3–4) | 29 (96.7) | |
| p-value | < 0.001 | |
Treatment-related toxicities were generally mild for most organ systems; however, hematologic toxicity was the most pronounced. The distribution of treatment-related toxicities according to CTCAE grade is presented in Supplementary Table S5. Infectious complications were significantly more frequent in patients with severe toxicity compared with those with mild or moderate toxicity (96.7% vs. 74.2%, p < 0.001) (Table 6).
Discussions
This study provides an integrated evaluation of population-level incidence trends and mortality patterns derived from national registry data, together with treatment-related toxicity analyzes from disease-specific pediatric cohorts treated at a national tertiary pediatric center. We observed significant increases in the incidence of osteosarcoma and neuroblastoma between 2016 and 2025, whereas ALL demonstrated a non-significant upward trend. Importantly, mortality remained statistically stable across all malignancies. At the clinical level, treatment-related toxicity was substantial, with predictors differing by disease: advanced disease stage was associated with severe toxicity in osteosarcoma, while risk group was the primary determinant of hematologic toxicity in ALL. In neuroblastoma, toxicity was most pronounced in patients with high-risk disease and intensive chemotherapy exposure. Together, these findings highlight a divergence between rising incidence and stable mortality, accompanied by substantial treatment-related morbidity.
The increasing incidence of osteosarcoma and neuroblastoma aligns with global reports describing heterogeneous but upward trends in certain pediatric cancers [10, 11]. Improved diagnostic capacity, enhanced cancer registry completeness, and demographic shifts may partially explain these patterns [12]. In Kazakhstan, substantial improvements in oncology registration and healthcare digitalization occurred during the early 2010s with the implementation of the national Electronic Registry of Cancer Patients [13]. This system improved case capture, reporting completeness, and centralized monitoring of oncology services. Therefore, part of the observed increase in incidence may reflect enhanced registration quality and improved case assurance rather than a true increase in disease occurrence alone. Additionally, advances in diagnostic infrastructure, wider access to specialized pediatric oncology services, and improved referral pathways may have contributed to increased detection of previously underdiagnosed cases [14–16].
Stable mortality despite rising incidence likely reflects advances in multimodal treatment and centralization of pediatric oncology services, trends observed in both high-income and transitioning health systems [6, 17]. However, stable mortality should not be interpreted as a reduction in disease burden. Rather, it may indicate improved survival accompanied by prolonged exposure to intensive therapies and their associated complications [6, 7, 18].
Our projection models suggest continued growth in incidence through 2030, with widening confidence intervals reflecting uncertainty inherent to time-series forecasting. Although males demonstrated a numerically higher projected incidence for osteosarcoma and neuroblastoma, sex-based differences were not statistically significant, suggesting broadly parallel trends across sexes. These projections underscore the need for health system preparedness, particularly in resource allocation, supportive care capacity, and long-term survivorship planning.
At the clinical level, toxicity patterns differ meaningfully by disease. In osteosarcoma, advanced disease stage was associated with severe toxicity, likely reflecting the greater treatment intensity required in patients with advanced disease. Chemotherapy regimen also influenced toxicity risk, highlighting the importance of protocol-specific risk stratification and individualized supportive care [19]. In ALL, treatment-related toxicity demonstrated a clear phase-dependent pattern, with the highest rates occurring during intensive blocks. Risk group was the primary predictor of hematologic toxicity and infectious complications. These findings may reflect protocol design, intensified monitoring, or differences in supportive care rather than reduced biological vulnerability [4]. Infectious complications were independently associated with interruption of maintenance therapy, emphasizing the clinical impact of toxicity on treatment continuity and potentially long-term outcomes.
In neuroblastoma, hematologic toxicity was concentrated in highly myelosuppressive chemotherapy blocks, consistent with the intensity of induction and consolidation regimens used in high-risk disease [8]. Toxicity was most pronounced in high-risk disease, reinforcing the central role of disease and treatment biology intensity in shaping toxicity burden.
The integration of epidemiologic forecasting with clinical toxicity data represents a key strength of this study. While survival gains in pediatric oncology are well documented [20], less attention has been paid to how increasing incidence and stable mortality interact with treatment-related morbidity at the population level. Our findings suggest that rising case numbers, even in the absence of increasing mortality, may substantially increase demand for supportive care services, infection management, transfusion support, and long-term survivorship monitoring.
Several limitations warrant consideration. First, the clinical cohort analyzes were based on retrospective data and may therefore be subject to selection and reporting biases. Second, the osteosarcoma cohort was relatively small, potentially limiting statistical power for subgroup analyses. Third, forecast estimates are inherently uncertain, as reflected by the widening confidence intervals over time. Fourth, although the study utilized national registry data, the findings may not be fully generalizable to other countries with different healthcare systems and epidemiological profiles. Additionally, the observed increase in incidence should be interpreted with caution. During the study period, Kazakhstan underwent substantial improvements in healthcare digitalization, cancer registration, and diagnostic infrastructure, including the implementation and expansion of the national Electronic Registry of Cancer Patients. Enhanced case assurance, improved reporting completeness, greater access to specialized oncology services, and advances in diagnostic technologies may have contributed to the increasing number of registered cases. Consequently, the available data do not allow us to distinguish whether the observed trends reflect a true increase in disease occurrence, improved detection and registration practices, or a combination of both factors. Furthermore, only three pediatric malignancies were included in the present analysis. Therefore, the findings should not be interpreted as representative of the entire spectrum of childhood cancers in Kazakhstan, and caution is warranted when extrapolating these results to pediatric oncology. Nevertheless, the integration of national registry-based epidemiological data with detailed clinical toxicity assessments provides a comprehensive perspective on the evolving burden and treatment-related challenges associated with osteosarcoma, neuroblastoma, and acute lymphoblastic leukemia in Kazakhstan.
Conclusion
Between 2016 and 2025, the incidence of osteosarcoma and neuroblastoma increased in Kazakhstan, while mortality remained statistically stable. Forecast models suggest that incidence will continue to rise through 2030. Despite stable survival outcomes, the burden associated with osteosarcoma, neuroblastoma, and acute lymphoblastic leukemia appears to be increasing. Advanced disease stage was associated with severe toxicity in osteosarcoma, while risk stratification was the primary determinant of hematologic toxicity in acute lymphoblastic leukemia. In neuroblastoma, toxicity was most pronounced among patients with high-risk disease receiving intensive therapy. These findings highlight that stable mortality does not indicate a reduced healthcare burden but rather an increasing need for strengthened pediatric oncology capacity, enhanced supportive care infrastructure, and long-term survivorship monitoring.
Supplementary Information
Acknowledgements
Not applicable.
Abbreviations
- APC
Annual percent change
- IRR
Incidence rate ratio
- CI
Confidence interval
- SE
Standard error
- ALL
Acute lymphoblastic leukemia
- WHO
World health organization, CTCAE - Common Terminology Criteria for Adverse Events
- AOR
Adjusted odds ratios
Authors’ contributions
GA conceptualized the study and supervised the research. All authors contributed to the study design, data collection, data analysis, and interpretation of the results. ZK, KK, and DK contributed to drafting the manuscript. AB, GT, and GA critically reviewed and edited the manuscript. All authors read and approved the final version of the manuscript.
Funding
The research was supported by program-targeted funding from the Ministry of Science and Higher Education of the Republic of Kazakhstan, «Development and implementation of a sustainable model of pediatric oncology care based on a precision approach in pediatric oncology», grant № BR28713341. JSC «Scientific Center of Pediatrics and Pediatric Surgery».
Data availability
The data and materials analyzed during the current study are available from the corresponding author upon reasonable request.
Declarations
Ethics approval and consent to participate
The study was conducted in accordance with the Declaration of Helsinki. Ethical approval was obtained from the Local Bioethics Committee of the Research Institute of Cardiology and Internal Diseases (Protocol No. 5/2025, dated October 7, 2025). Due to the retrospective design of the study and the use of anonymized data, the requirement for informed consent was waived.
Consent for publication
Not applicable, as the study used anonymized retrospective data and did not include identifiable patient information.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.World Health Organization. Childhood cancer. Published February 4. 2025. https://www.who.int/news-room/fact-sheets/detail/cancer-in-children
- 2.Mussina K, Kuanova B, Syssoyev D, et al. Epidemiology of pediatric hematological malignancies in Kazakhstan: Data from Unified National Electronic Healthcare System 2014–2021. Eur J Pediatr. 2024;183(4):1683–91. 10.1007/s00431-023-05412-3. [DOI] [PubMed] [Google Scholar]
- 3.World Health Organization. Cancer country profile 2020: Kazakhstan. Geneva, Switzerland: World Health Organization; 2020. https://www.who.int/publications/m/item/cancer-kaz-2020. [Google Scholar]
- 4.Hunger SP, Mullighan CG. Acute Lymphoblastic Leukemia in Children. N Engl J Med. 2015;373(16):1541–52. 10.1056/NEJMra1400972. [DOI] [PubMed] [Google Scholar]
- 5.Pui CH, Yang JJ, Hunger SP, Pieters R, Schrappe M, Biondi A, Vora A, Baruchel A, Silverman LB, Schmiegelow K, Escherich G, Horibe K, Benoit YC, Izraeli S, Yeoh AE, Liang DC, Downing JR, Evans WE, Relling MV, Mullighan CG. Childhood Acute Lymphoblastic Leukemia: Progress Through Collaboration. J Clin Oncol. 2015;33(27):2938–48. 10.1200/JCO.2014.59.1636 . Epub 2015 Aug 24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Angeles Vázquez López M. Childhood cancer survivors: current challenges. Pediatr (Engl Ed). 2024;100(5):363–75. 10.1016/j.anpede.2024.04.013. [DOI] [PubMed] [Google Scholar]
- 7.Martos M, George N, Arreguín-González FE, et al. Childhood Cancer Survivorship Globally: A Systematic Review. JAMA Oncol Published online August. 2025;14. 10.1001/jamaoncol.2025.2506. [DOI] [PubMed]
- 8.Irwin MS, Park JR. Neuroblastoma: paradigm for precision medicine. Pediatr Clin North Am. 2015;62(1):225–56. 10.1016/j.pcl.2014.09.015. [DOI] [PubMed] [Google Scholar]
- 9.Mirabello L, Troisi RJ, Savage SA. Osteosarcoma incidence and survival rates from 1973 to 2004: data from the Surveillance, Epidemiology, and End Results Program. Cancer. 2009;115(7):1531–43. 10.1002/cncr.24121. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Sultan I, Alfaar AS, Sultan Y, Salman Z, Qaddoumi I. Trends in childhood cancer: Incidence and survival analysis over 45 years of SEER data. PLoS ONE. 2025;20(1):e0314592. 10.1371/journal.pone.0314592. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Steliarova-Foucher E, Colombet M, Ries LAG, et al. International incidence of childhood cancer, 2001–10: a population-based registry study. Lancet Oncol. 2017;18(6):719–31. 10.1016/S1470-2045(17)30186-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Bray F, Parkin DM. Evaluation of data quality in the cancer registry. Eur J Cancer. 2009;45(5):747–55. [DOI] [PubMed] [Google Scholar]
- 13.Republican Center for Electronic Health. Electronic Registry of Cancer Patients (EROP). Ministry of Healthcare of the Republic of Kazakhstan. Available from: https://rcez.kz/informationsystems. Accessed June 14, 2026.
- 14.Kosherbayeva L, Hailey D, Kurakbaev K, et al. Implementation of health technology assessment work in a hospital in Kazakhstan. Int J Technol Assess Health Care. 2016;32(1–2):78–80. 10.1017/S0266462316000076. [DOI] [PubMed] [Google Scholar]
- 15.Kosherbayeva L, Hailey D, Kurakbaev K, et al. A process of prioritizing topics for health technology assessment in Kazakhstan. Int J Technol Assess Health Care. 2016;32(3):147–51. 10.1017/S0266462316000222. [DOI] [PubMed] [Google Scholar]
- 16.Government of the Republic of Kazakhstan. Resolution No. 356 of May 2, 2007. On Certain Issues of Implementation of the Project ‘Construction of 100 Schools and 100 Hospitals Based on Public-Private Partnership’. Astana: Government of the Republic of Kazakhstan. 2007. Available from: https://adilet.zan.kz. [Cited 2026 Jun 14].
- 17.Gatta G, Botta L, Rossi S, et al. Childhood cancer survival in Europe 1999–2007: results of EUROCARE-5. Lancet Oncol. 2014;15(1):35–47. [DOI] [PubMed] [Google Scholar]
- 18.Smith MA, Seibel NL, Altekruse SF, et al. Outcomes for children and adolescents with cancer: challenges for the twenty-first century. J Clin Oncol. 2010;28(15):2625–34. 10.1200/JCO.2009.27.0421. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Mirabello L, Troisi RJ, Savage SA. Osteosarcoma incidence and survival rates. Cancer. 2009;115(7):1531–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Allemani C, Matsuda T, Di Carlo V, et al. Global surveillance of trends in cancer survival 2000–14 (CONCORD-3). Lancet. 2018;391(10125):1023–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data and materials analyzed during the current study are available from the corresponding author upon reasonable request.