ABSTRACT
Objective
Neuroblastoma (NB) is a common childhood tumour, affecting telomerase enzyme activity. Expression of telomerase reverse transcriptase (TERT) protein is crucial for the fuınctioning of telomerase activity, but its association with risk stratification and prognosis is unclear. The adenosine triphosphate-dependent helicase alpha-thalassemia/mental retardation x-linked (ATRX) protein, a chromatin remodeling protein, accumulates H3.3 histone variants. The study aimed to assess the correlation between expression levels of ATRX and TERT with the NB risk group and its prognosis.
Method
Immunohistochemical expressions of TERT and ATRX proteins in tumour tissue samples of 54 NB cases at different stages and risk groups were evaluated.
Results
Immunohistochemical expression rates of TERT and ATRX proteins in tissues were 55.8% and 61.2%, respectively, with ATRX positively expressed at a rate of 50%, and 69% in early and in advanced stages of NB, and rates of 67.7%, and 50% in high and low-risk groups, respectively. TERT expression varies in early and advanced stages of NB, with higher levels in high-risk groups. ATRX expression is significantly higher in NB patients with Neuroblastoma myc (NMYC) gene amplification.
Conclusion
High expression of ATRX in NB patients with NMYC gene amplification suggests that ATRX may be used as a potential immunohistochemical prognostic marker in NB patients.
INTRODUCTION
Neuroblastoma (NB) is the most common extracranial solid tumour in children. This tumour, which is frequently seen in children younger than two years of age, is defined in 90% of children younger than five years of age(1). NB may show spontaneous regression or benign transformation to ganglinouroma especially in patients under 1 year of age. In patients over 1 year of age, the disease has a more aggressive course(1). Genetic changes play an important role in the prognosis and treatment of the disease(2). In the risk classification made for the treatment of the disease, genetic mutations and chromosomal changes in the patient are also examined in addition to the patient’s age, tumour stage and histology(3, 4). MYCN oncogene amplification is an amplification observed in 20-30% of NB patients and is associated with poor prognosis(5) In advanced NB patients with MYCN amplification, event-free and overall survival (OS) rates are reported to be considerably lower compared to the patients without(6). According to the The Turkish Pediatric Oncology Group (TPOG) - NB 2009 Protocol study conducted between 2009 and 2020, 70% of the patients diagnosed in Türkiye had advanced disease and 59% of these patients were in the high risk group according to the NB diagnosis and treatment protocol(4). The tumour suppressor gene alpha-thalassemia/mental retardation X-linked (ATRX) blocks DNA replication and transcription by taking part in a chromatin remodelling whose main function is the accumulation of histone variant H3.3(7). ATRX mutations are commonly found in glioma and are associated with the development of alternative telomere lengthening (ALT), a non-telomerase-dependent telomere lengthening mechanism(8). In addition to its known roles, it influences various cellular processes associated with epigenetic regulation(9). The loss of ATRX can arise through gene mutations, deletions, or chromosomal rearrangements such as gene fusions. The ALT phenotype is frequently linked to distinct molecular changes, including amplification of the platelet-derived growth factor receptor-alpha and mutations in the tumour suppressor gene tumor tumor protein p53. In most NB cells, upregulation of telomerase activity or activation of the ALT pathway results in activation of telomere maintenance mechanisms. Activation of the ALT pathway is mostly caused by mutations in the ATRX gene. As a result of this activation, telomeres elongate leading to carcinogenesis. ATRX mutations are observed especially in NB patients older than 18 months and show a positive correlation with age(7). Telomerase is a key enzyme involved in regulating cell proliferation and tumorigenesis. It maintains chromosomal integrity by adding hexameric sequences to the ends of chromosomes, thereby preventing telomere shortening and the onset of cellular senescence. The catalytic subunit of human telomerase, known as hTERT, serves as the primary rate-limiting component for telomerase activity(10). This enzyme is active in over 90% of human malignancies. In cancer, telomerase reverse transcriptase (TERT) expression can be elevated through multiple mechanisms, such as gene amplifications, promoter mutations, and chromosomal rearrangements. Notably, promoter rearrangements of TERT have been identified in high-risk NB cases(11). In NB, determination of prognostic markers that can provide information about the risk groups and prognosis of patients is very important both for the development of targeted therapies and for obtaining better treatment results. Some molecules that may play a role in the prognosis of NB or treatment response continue to be revealed(12-17). In this study, we investigated TERT and ATRX protein expressions in tumour tissues of early-and late- stage NB patients from different NB risk groups and evaluated whether or not their expression levels could be differentiating markers in terms of aggressiveness in early- and late- stages of NB by revealing their relationship with NB risk groups and prognosis.
MATERIALS and METHODS
Experimental Groups
Permission was obtained from Dokuz Eylül University (DEU) Non-Interventional Research Ethics Committee for the conduction of this research study (decision no: 2023/14-10, dated: 03.05.2023). Signed consent forms were obtained from the patients during the sample collection and storage stages. The study was carried out using tumour paraffin tissue samples of 53 NB patients submitted to DEU Institute of Oncology Department of Basic Oncology within the scope of TPOG-2020 protocol. Tissue samples of NB patients whose molecular analyses were performed and known to be in low and advanced stages of NB were used in the study(12, 18).
Study population of 53 people consisted of 22 girls and 31 boys (Table 1). In NB, stage L1 refers to localised tumours confined within a single anatomical compartment without any image-defined risk factors (IDRFs). Stage L2 is characterised by regional tumours with the presence of IDRFs. The tumour may extend into different compartments on the ipsilateral side of the body. Stage M includes cases with distant metastatic spread. Involvement of the bone, liver, distant lymph nodes, or pleural/abdominal effusion containing malignant cells outside the primary site indicates metastatic disease. Finally, stage MS represents a special form of metastatic disease that occurs only in patients younger than 18 months. It is defined by limited involvement of the skin, liver, and/or bone marrow (less than 10% of total marrow cellularity)(1). In NB molecular analyses, real-time polymerase chain reaction (RT-PCR) analyses were performed to detect NMYC, 11q, 1p and 17q. Out of 53 patients, 22 were in the high-risk, 9 in the intermediate-risk and 18 in the low-risk group.
MYCN amplification status was defined based on RT-PCR results, with samples showing greater than a 10-fold increase were classified as positive, and those below this threshold as negative. As part of the TPOG study, NB Formalin-Fixed Paraffin-Embedded tissue samples at various disease stages were randomly selected from the archival collection of the Departments of Paediatric Oncology and Basic Oncology at Dokuz Eylül University. Tumour sections were cut from paraffin-embedded blocks and mounted on adhesive slides for immunohistochemical (IHC) analysis of ATRX and TERT protein expressions. According to the TPOG-2020 classification system, the patients were categorised into high-risk (n=22), intermediate-risk (n=9), and low-risk (n=18) groups(4, 19).
Antibodies
In this study polyclonal human ATRX and TERT antibodies (Bioss, Inc. 500 West Cummings ParkSuite 6500 Woburn, MA, USA) were used, along with a secondary antibody obtained from Ventana. All antibodies were stored in accordance with the manufacturers’ recommendations, and appropriate dilution ratios were observed during application. Optimal dilutions were determined through control staining, with both ATRX and TERT antibodies used at a 1:100 dilution. Paraffin-embedded tissue blocks of tumour specimens of NB patients were selected to represent all three risk categories to ensure comprehensive analysis across the disease spectrum.
IHC Staining
Tissue sections were incubated overnight at 60 °C in an oven prior to application of the staining procedure(18). Following incubation, the slides underwent deparaffinization in xylene for one hour, and were subsequently rehydrated using a graded series of alcohol solutions. Antigen retrieval was achieved by heating the slides in citrate buffer using a microwave. After treatment with hydrogen peroxide and appropriate washing steps, human-specific primary Immunoglobulin G antibodies were applied, based on prior optimization. Subsequently, a multimer Horseradish Peroxidase-conjugated secondary antibody was added and allowed to incubate. In the final staining step, diaminobenzidine combined with hydrogen peroxide was used to catalyze the chromogenic reaction. The nuclei of the tumour cells were then counterstained with haematoxylin, and the slides were passed through an ascending alcohol series before being cleared in xylene. The prepared slides were examined under a light microscope (Olympus BX50). ATRX expression was detected in the nuclei, whereas TERT protein exhibited both nuclear and cytoplasmic localization. To evaluate staining intensities of ATRX and TERT, five randomly selected fields per section were analyzed for each sample. Tumour cell staining intensities were classified as follows: 0= no staining; 1= weak; 2= moderate; 3= strong. Additionally, the proportion of stained cells for ATRX and TERT was scored as 0= none; 1=0-20%; 2=21-50%; 3=51-80%; 4=81-100%. Given the strong correlation between staining intensities and area scores, the intensity score was utilized for statistical analysis in this study(20, 21).
Determination of NMYC and 11q by RT-PCR and Flow Cytometry in Molecular Evaluation
Paraffin-embedded NB tissue samples were used for DNA isolation, and concentrations were quantified fluorimetrically using a Qubit® fluorometer. RT-PCR was performed to assess N-MYCN amplification and 11q23 deletion. Threshold cycle values were determined for target and reference genes in patient and control DNA samples. Copy number alterations were detected using specific primers and TaqMan probes: for N-MYCN, primers 5’-GTGCTCTTCCAATTCTCGCCT-3’ and 5’-GATGGCCTAGAGGAGGGCT-3’ with a 6-carboxyfluorescein (FAM)-labeled TaqMan probe; for 11q23 (ARCN1 gene), primers 5’-ATCTGGAGGCAGCACAGCT-3’ and 5’ TACACTGGATTATACCCTGGCTGG-3’ with a FAM-labeled probe. PCR reactions were performed in eight replicates on a Roche Nano RT-PCR system. Relative quantification was calculated using the ΔΔCT method with healthy reference DNA as a calibrator, and findings were further validated via absolute quantitation using reference DNA standards (Table 2)(12, 18).
Multivariate Survival Analysis
The patients included in the study were clinically categorised according to disease stage and risk classification. Patients were also categorised according to the molecular, and histological characteristics of the tumour, and age at diagnosis. The risk assessment and disease stage of the patients were considered to be confounding variables. Therefore, multivariate Cox regression survival analysis including the previously mentioned confounding factors was performed to confirm our findings. Cox regression analysis using different models showed that ATRX and TERT did not differ in univariate survival analysis. Again, no statistically significant difference was found between these parameters in multivariate Cox regression analysis. However, in the models created in Cox regression analysis, a significant relationship was detected between N-MYC Amp in NB patients and event-free survival (EFS) and OS rates (Table 3).
Statistical Analysis
We used Fisher’s exact test to assess the relationship between categorical variables and ATRX and TERT expression patterns. The Independent Samples t-test was used to compare mean EFS and OS times between groups. Data were presented as mean ± standard deviation and the number of observations. EFS was defined as the time to the emergence of recurrence, secondary malignancy, or death. OS analysis included the interval from study registration to death or last follow-up. Survival curves were generated using the Kaplan-Meier method, and differences were tested using the log-rank test. All statistical analyses were conducted using IBM SPSS Statistics Version 29 (IBM Corp., USA), with p-values <0.05 considered statistically significant.
RESULTS
Data on the Clinicopathology of NB Patients
In this study, tissue samples from 22 female, and 31 male patients diagnosed with NB were analysed. The patients’ ages ranged from 1 month to 11 years. Staging was performed in accordance with the criteria established by the Turkish Paediatric Oncology Group (2020) and the International Neuroblastoma Staging System. Risk stratification was based on the International Neuroblastoma Risk Group Staging System (INRGSS). According to the applied classification system, patients were categorised in the high (n=22), intermediate (n=9), and low-risk (n=18) groups, respectively (Table 1).
Risk Cassifications of Patients
Twenty-two-high, nine intermediate, and eighteen low-risk patients were analysed in this study. Both molecular (MYCN amplification, 11q23 loss and DNA index) and clinical (age of the patient, stage and histopathology of the tumour) data of patients were evaluated for the risk classification (Table 4).
Association of Age with Molecular Alterations and Survival Outcomes
To assess the prognostic impact of age, patients were stratified into two groups based on the well-established prognostic cut-off of 5 years of age (<5 years, n=38; ≥5 years, n=15). Loss of ATRX nuclear expression (negative staining) was significantly more frequently detected in patients aged 5 years or older compared to younger patients (46.7% vs. 15.8%, p=0.021). Similarly, a non-significant trend towards higher TERT expression was observed in the older age group (73.3% vs. 50%, p=0.13). Kaplan-Meier survival analysis revealed that patients ≥5 years of age had significantly worse OS compared to those <5 years (5-year OS: 40% vs. 78%, p=0.009 by log-rank test). EFS also followed a similar though non-significant, trend (5-year EFS: 33% vs. 63%, p=0.058). In a multivariate Cox regression model adjusted for MYCN amplification status and INRGSS risk groups, age ≥5 years remained an independent predictor of poorer OS [hazard ratio (HR): 3.2, 95%, confidence interval (CI): 1.1-9.3, p=0.032].
Expression of TERT and ATRX in Tissue Samples of Patients
In tissue samples, TERT and ATRX were expressed in 55.8% and 61.2% of cases, respectively. In NB, ATRX expression increased with disease stage, and observed in 50% of early-stage and 69% of advanced-stage tumours. Similarly, ATRX expression was higher in high-risk patients (67.7%) compared to low-risk patients (50%). TERT was expressed at a rate of 55% in the early stage and 62.1% in the advanced stage. TERT was expressed at a rate of 58.1% in high-risk groups and 55.6% in low-risk groups. ATRX and TERT alterations were common in MYCN-amplified tumours (ATRX 86.7%, 13/15; 95% CI: 62.1-96.3; TERT 73.3%, 11/15; 95% CI: 48.0-89.1). In MYCN-non-amplified cases, the corresponding frequencies were 61.2% (23/38; 95% CI: 44.7-74.4) and 55.8% (21/38; 95% CI: 39.7-69.9), respectively. No significant differences were observed between groups in terms of expression rates of these risk predictors (ATRX: p=0.10; TERT: p=0.35). ATRX expression was found to be significantly higher in NB patients with MYCN amplification compared to patients without (p=0.027) (Figure 1 and 2).
Association Between ATRX and TERT Expressions and EFS and OS in High and Low Risk Groups in NB
The prognostic value of ATRX and TERT expressions in NB was assessed by stratifying patients into “high” and “low” expression groups. Associations between gene expression and OS, EFS, event occurrence, and risk classification were evaluated using Kaplan-Meier analysis, with group comparisons conducted via the log-rank test. No significant differences were observed in terms of OS or EFS between ATRX-defined groups (OS: p=0.294; EFS: p=0.337) and TERT-defined groups (OS: p=0.693; EFS: p=0.740). Additionally, expressions of ATRX and TERT showed no significant correlation with event occurrence or risk classification (all p>0.05). Kaplan-Meier curves indicated substantial overlap in survival probabilities across expression-defined groups.
These results suggest that expressions of ATRX and TERT per se are not independent prognostic indicators in this cohort and are insufficient to predict survival outcomes. Their potential clinical utility may require their integration with additional molecular markers in a multivariate analytical framework. The corresponding survival curves are shown in Figure 3.
DISCUSSION
Ongoing advancements in NB research have contributed to a modest improvement in patient prognosis over recent years, with a corresponding increase in the 5-year survival rates. While patients with low-risk NB have shown relatively high cure rates, outcomes for patients with high-risk NB have remained largely unchanged. As such, elucidating the molecular characteristics and genetic alterations underlying NB is critical for developing strategies that enable early diagnosis and implementation of targeted therapeutic approaches(22). Notably, ATRX mutations and the loss of nuclear protein expression have been observed more commonly in patients over the age of 12 who present with stage 4 disease. In a study by Cheung et al.(23), ATRX gene deletions were identified in 43% of older adolescents with advanced NB (>12 years) and in 11% of paediatric patients aged 5-12 years. Similarly, in our study, OS rates were significantly worse in the ≥5 age group, and age was found to be an independent risk factor in multivariate analysis, independent of other prognostic parameters such as MYCN status and risk group. These findings suggest that age is associated not only with clinical stage and molecular alterations, but also with tumour biology, including loss of ATRX mutations and telomere mechanisms. Therefore, age is a strong prognostic marker in NB that reflects molecular subgroups beyond classical risk classifications.
According to Clusters of Orthologous Genes database data, survival was significantly preserved in patients assigned to less intensive treatment due to the change in the age threshold from 12 to 18 months. This finding supports that biologically “more favourable” tumours in younger age groups do not always require intensive treatment and that age can be safely used as a variable in intensification of treatment. In this study, the better prognosis of the <5 age group also demonstrates that the younger age group has biologically more “favourable” tumours and can achieve better survival without requiring intensive treatment(24).
Data from a separate study related to the ≥5 year- and adolescent/young adult categories have shown a progressive, and significant drop in survival rates with advancing age. adolescent/young and adult patients have demonstrated poorer OS compared to children in Surveillance, Epidemiology, and End Results -based studies, even when detected earlier which suggests both stage of NB and biological differences as determinants of survival. Ten-year OS has been observed to decrease to around 19% in adult-onset NB series. In keeping with previous research, our findings in this study have also demonstrated a notable drop of 40% in OS rates in the group of children aged ≥5 years. The trend is similar, even though this figure is better than the rates indicated for adults and adolescents in the literature(25).
In NB, telomere length has been investigated as a significant prognostic indicator, with evidence indicating that shorter telomeres are associated with a more favourable clinical outcome, whereas longer or unchanged telomere lengths correlate with poorer prognoses(20). More recently, activation of telomerase due to genomic rearrangements near the TERT gene locus has been identified in NB, delineating a subgroup of high-risk tumours characterized by exceptionally poor survival outcomes(13, 26). These investigations have demonstrated that TERT rearrangements are linked to elevated levels of TERT mRNA and enhanced telomerase enzymatic activity. In the study conducted by Lee et al.(20), TERT expression was assessed through IHC staining, revealing an inverse, but statistically insignificant association between TERT expression and patient survival. Similarly, in the current study, although a marked increase in TERT expression was observed in tissue samples of patients, still the inverse relationship with survival lacked statistical significance. Additionally, one of the earliest reports identifying ATRX involvement in NB revealed ATRX mutations in 44% of metastatic NB cases among adolescents and young adults, whereas such mutations were absent in tumour tissues obtained from infants with metastatic disease(23). Children whose tumours harbored ATRX mutations were generally older than five years or exhibited a more indolent or chronic disease progression. In the same investigation, ATRX mutations were found to occur independently of MYCN amplification and were linked with nuclear loss of ATRX protein, telomere lengthening, and the activation of the alternative lengthening of telomeres (ALT) pathway(23). Our findings also support this trend. Indeed OS was significantly worse in the ≥5 age group and remained an independent risk factor in the multivariate model, independent of age, MYCN status, and INRGSS risk group (HR: 3.2; 95% CI: 1.1-9.3; p=0.032). This result suggests that age acts as a higher-level marker that captures biological differences on the ATRX/TERT axis in addition to the classic risk classification.
Another study indicated that the majority of high-risk NB tumours exhibit either TERT rearrangements, MYCN amplification, or ATRX mutations —alterations that collectively promote telomere elongation and provide a molecular basis for defining this subtype of NB(26). Conversely, low-risk tumours are typically devoid of these genomic alterations and exhibit low TERT expression, which may reflect an inability to achieve unlimited proliferative capacity. The most aggressive subtype of NB has been associated with telomerase activation resulting from either TERT rearrangements or MYCN amplification. With ongoing advances in telomerase inhibitor development, these findings may offer a promising therapeutic avenue for treating the most lethal forms of this paediatric malignancy. Furthermore, three recent reports have shown that ATRX mutations frequently cause loss of protein expression in NB and are more prevalent in older patients and those diagnosed with stage IV disease(23, 27, 28). In our cohort, MYCN amplification was also found to be significantly associated with EFS and OS in multivariate analysis, consistent with the rekevant literature data. Since ATRX and TERT expressions per se fail to predict survival these biomarkers should be evaluated together with multiple parameters rather than independently in clinical practice.
Only 4 out of 53 patients (7.5%) in our dataset had both MYCN amplification and 11q deletion, indicating a trend of reciprocal exclusivity; however, this negative correlation was not statistically significant (p=0.145). This finding is directionally consistent with the well-established genomic landscape of NB, where these two alterations are known to rarely coincide and define distinct molecular subtypes demonstrating aggressive clinical behavior(29, 30). For instance, a comprehensive genomic study by Molenaar et al.(28) in 2012 on 87 NB cases found a strong pattern of mutual exclusivity between MYCN amplification and 11q loss, a hallmark that helps stratify high-risk disease. Similarly, the large-scale study performed by Pugh et al.(27) in 2013, which analyzed the genetic landscape of 240 high-risk NBs, confirmed that 11q deletion and MYCN amplification are significantly and mutually exclusive events (p<0.001), underscoring their roles in alternative pathways of oncogenesis. Limited sample size of our study compared to large cohort studies may reduce statistical power of our conclusions. Nevertheless, the very low frequency of co-occurrence observed among these parameters reinforces the concept that these are separate oncogenic drivers.
CONCLUSION
The present study revealed a relationship between expression levels of ATRX and TERT and both the NB risk classification and clinical prognosis. Interestingly, ATRX expression was found to be higher in NB patients, especially in those with MYCN amplification, suggesting its potential utility as an IHC prognostic marker. Considering their biological significance, ATRX and TERT emerge as important molecular candidates that may enhance our understanding of NB pathology. Nonetheless, their exact mechanisms of action in NB are yet to be fully elucidated. To clarify their roles, future research should be performed with larger patient cohorts and should focus on in-depth mechanistic investigations, supported by validation in clinical tissue samples.
