Background

Neuroblastoma is the most common extracranial malignancy in children with remarkable heterogeneity in clinical behaviors. Neuroblastoma accounts for 10%–15% of all pediatric cancer-related deaths [1, 2]. Although some of the tumors can regress spontaneously, most patients’ diseasees progress aggressively despite multimodality treatment, which may include chemotherapy, radiotherapy, surgical resection, hematopoietic stem cell transplantation, and immunotherapy [3, 4]. In the clinic, half of neuroblastoma patients present with metastasis at diagnosis, and less than 50% survive 5 years [5, 6].

Recently, many prognostic variables have been used for risk stratification to predict the survival of patients with neuroblastoma. Among them, genetic features such as the status of the v-myc avian myelocytomatosis viral oncogene neuroblastoma derived homolog (MYCN) gene [7], 11q23 allele status [8], and tumor ploidy [9, 10] are the most significant and clinically relevant factors. However, for pediatric patients with metastatic neuroblastoma who need chemotherapy prior to surgery, timely identification of genetic aberrations is not possible because of the difficulty in obtaining tumor biopsies. In addition, the results of genetic aberration identification may be inaccurate after the patients have received chemotherapy, leading to insufficient dosage and period of treatment [11, 12].

For more than two decades, a number of genetic aberrations closely associated with outcome have been identified in neuroblastoma; for the MYCN gene, these include 17q gain, loss of 3p, and 11q deletion [13]. Among them, 11q deletion is the most frequent, occurring in approximately 40% of primary neuroblastomas [14], and is associated with poor prognosis [15, 16]. However, the clinical significance of 11q deletion is unclear in neuroblastoma metastatic to bone marrow. What’s more, 11q deletion is not routinely examined when genome-wide methods are used to evaluate genetic aberrations [17]. 11q23 is the most commonly deleted 11q region, and interphase fluorescence in situ hybridization (FISH) has been previously confirmed as a reliable method for detecting MYCN status in bone marrow metastases of neuroblastoma [18, 19]. Therefore, in this study, we used FISH to simultaneously detect MYCN and 11q23 deletion in bone marrow cells in order to evaluate the biological characteristics and survival associated with these two markers.

Materials and methods

Patients and treatment protocols

Consecutive children with neuroblastoma newly diagnosed at the Hematology Oncology Center, Beijing Children’s Hospital, Capital Medical University between January 2015 and December 2016 were selected. All patients were diagnosed according to the International Neuroblastoma Staging System [20]. The criteria for patient inclusion were ≥ 20% neuroblastoma cells in the bone marrow samples and treatment according to the BCH-NB-2007 protocol (based on the Hong Kong neuroblastoma protocol) [21].

The BCH-NB-2007 protocol for newly diagnosed high-risk neuroblastoma was comprised of 7 cycles of intensive chemotherapy: CAV regimen [cyclophosphamide (70 mg/kg) on days 1–2, adriamycin (25 mg/m2) on days 1–3, and vincristine (0.033 mg/kg)] on days 1–3 for cycles 1, 2, 4, and 6; CVP regimen [cisplatin (50 mg/m2) on days 1–4 and VP16 (200 mg/m2) on days 1–3] for cycles 3, 5, and 7. Primary tumor resection was performed after 4 cycles of chemotherapy; peripheral blood stem cells (PBSCs) were then harvested for possible autologous bone marrow transplantation. Autologous PBSC transplantation (PBSCT) was performed after 7 cycles of chemotherapy. The patients received local irradiation (20–25 Gy) at the site of primary tumor as indicated [22]. At 4–5 weeks after autologous PBSCT, they received maintenance chemotherapy with cis-retinoic acid (160 mg/m2) for 14 days alternating with 14 days off for 9–12 months.

The treatment protocol for low-/intermediate-risk neuroblastoma was comprised of 2–3 cycles of CBVP regimen [carboplatin (200 mg/m2) and etoposide (150 mg/m2)] and CADO regimen [vincristine (1.5 mg/m2) and adriamycin (25 mg/m2)] used alternately. Primary tumor resection was usually performed after chemotherapy. If surgery was performed before chemotherapy, 2–3 cycles of chemotherapy for low-risk patients and 4–6 cycles for intermediate-risk patients were needed.

All patients were treated and evaluated according to the BCH-NB-2007 protocol. After the treatment, the patients were followed-up every 3 months mainly by outpatient follow-up visits (bone marrow examination and computed tomography or magnetic resonance imaging) and telephone interview. The final date for data collection was August 31, 2018. This study was approved by the Beijing Children’s Hospital Institutional Ethics Committee. Informed consent was obtained from the parents or their guardians in accordance with the Declaration of Helsinki.

Bone marrow sample processing

Bone marrow samples of all patients obtained by bone marrow puncture were collected in heparin-coated tubes at diagnosis. The sample was immediately shaken gently after collection to avoid bone marrow agglutination. Then, the cells were cultured in 8 mL RPMI-1640 (Hyclone, Beijing, China) and 2 mL fetal bovine serum medium (Invitrogen, Carlsbad, CA, USA) for 24 h. The cell density was (1 − 3) × 106/mL, and the total amount of cells was about (1 − 3) × 107. Colchicine was added at 2 h before termination of the culture. The cells were then treated with 0.075 mol/L KCl for 30 min and fixed twice in a 3:1 mixture of methanol:acetic acid. Finally, the cell suspension was stored in a fresh fixative solution at 4 °C. Bone marrow biopsies were prepared on wax blocks for morphological examination.

FISH analysis of bone marrow cells

The FISH technique was performed as previously reported [17]. The statuses of MYCN and 11q were determined with DNA probes from Vysis [N-MYC(2p24)/CEP2(2p11.1-q11.1) Dual Color Probe and LSI MLL Dual Color, Break Apart Rearrangement Probe] (cat. No. 7J72-01 and 8L57-20, Abbott Laboratories, Abbott Park, IL, USA). A Leica DM6000B microscope (Leica Microsystems GmbH, Wetzlar, Germany) was used to take fluorescence images. According to the recommendations of the European Neuroblastoma Quality Assessment group [23, 24], MYCN amplification was defined as a > fourfold increase of MYCN signals in relation to the number of chromosome 2, and 11q23 deletion was defined as only one fusion signal displayed.

Karyotype analysis of bone marrow cells

Chromosome karyotype analysis of bone marrow cells was performed on G-banded preparations. At least 20 metaphases were analyzed in detail. The karyotype results were interpreted according to the International System for Human Cytogenomic Nomenclature guidelines (2016) [25]. The clones containing at least 2 cells with the same additional chromosome or structural abnormality or at least 3 cells with the same missing chromosome were defined as abnormal clones.

Morphologic analysis and diagnostic biomarker detection

Microscopic examination of bone marrow biopsies to determine the presence of neuroblastoma cells was done by at least 2 independent laboratory experts. Serum tumor markers such as lactate dehydrogenase (LDH) and neuron specific enolase (NSE) levels were detected by radioimmunoassay and full-automatic biochemical analysis at the time of diagnosis in all patients.

Statistical analysis

All statistical analyses were performed using the SPSS software, version 16.0 (SPSS Inc., Chicago, IL, USA). Comparisons of patients with or without 11q23 deletion and MYCN amplification and the association of clinical characteristics were evaluated using the χ2 test. Event-free survival (EFS) was defined from the date of diagnosis to the date of one of the following events: progression, relapse, death, or last contact with patients in continuous complete remission (CR). Overall survival (OS) was defined from the date of diagnosis to the date of death due to any reason or the last contact with patients in continuous CR. Kaplan–Meier survival analysis and log-rank test were used to determine the differences in EFS and OS in all patients by MYCN and 11q23 statuses. Multivariate backward Cox regression analysis model was built which included only those variables reaching a P value < 0.05 in the respective univariate analysis. All tests with a P < 0.05 in two-sided distributions were considered statistically significant.

Results

Clinical characteristics of 101 patients with bone marrow metastatic neuroblastoma

Among the 161 patients with newly diagnosed neuroblastoma, 45 had no bone marrow metastasis, 13 gave up treatment after diagnosis, and 2 were lost to follow-up. A total of 101 patients with stage IV neuroblastoma were thereby analyzed, including 57 males and 44 females in a ratio of 1.3:1. The patients were aged 7–105 months (median, 36 months), and 85 (84.2%) were older than 18 months. The median white blood cell (WBC) count was 5.6 × 109/L (range 1.7–42.5 × 109/L) with 60 (59.4%) patients having normal WBC. Only 8 (7.9%) patients had hemoglobin > 120 g/L. The rest had different degrees of anemia, but most cases were mild. The median hemoglobin level was 94 g/L (range 58–135 g/L). The amount of platelets ranged from 37 × 109/L to 622 × 109/L (median, 275 × 109/L). Almost all patients (92.1%) were grouped as having high-risk neuroblastoma. The primary tumor sites were the abdomen for 88 patients, thorax for 12 patients, and neck for 1 patient. 11q23 deletion was seen in 40 (39.6%) patients, and MYCN amplification in 12 (11.9%). In terms of clinical serum tumor markers, only 22 (21.8%) patients had LDH levels ≥ 1500 IU/L, and 41 (40.6%) had NSE levels ≥ 370 ng/mL.

Each patient received a total of 7 cycles of chemotherapy except for those who died during therapy. During follow-up, 39 (38.6%) patients died, and 13 (12.9%) suffered from tumor relapse or progression; the other 49 (48.5%) had stable disease. Among the patients who died, 11 died of tumor progression during intensive chemotherapy, 24 died of tumor progression during maintenance treatment, and 4 died of tumor recurrence after the completion of chemotherapy. Of note, 4 of the 24 patients who died during maintenance treatment had intracranial metastasis. The main clinical characteristics of the 101 patients are shown in Table 1.

Table 1 Baseline characteristics of the 101 patients with bone marrow metastatic neuroblastoma
Full size table

Genetic markers of MYCN and 11q23 in bone marrow cells

In all patients, interphase FISH detected the MYCN gene and 11q23 status of bone marrow cells. Chromosome karyotype analysis was performed in 55 patients. MYCN amplification was seen in 12 (11.9%) patients, with > 30 gene copies/cell (Fig. 1a). MYCN gain was seen in 21 (20.8%) patients, with 2–9 signal copies. Among the 101 patients, 40 (39.6%) had 11q23 deletion (Fig. 1b). Only 2 (2.0%) patients had both MYCN amplification and 11q23 deletion. MYCN amplification and 11q23 deletion were not associated with higher percentages of tumor cells in bone marrow or more severe hematopoietic suppression (both P > 0.05).

Fig. 1
figure 1

Typical fluorescence in situ hybridization (FISH) images of bone marrow cells from patients with bone marrow metastatic neuroblastoma. a The status of v-myc avian myelocytomatosis viral oncogene neuroblastoma derived homolog (MYCN) amplification was detected using a dual-color probe. Green signals represent the specific probe for MYCN, and red signals stand for centromeric chromosome 2 probes. MYCN signals show more than 10 copies within the nuclei. b The status of 11q23 deletion was also detected using a dual-color probe. Normal cells show a two orange/green fusion signal pattern. In a cell with 11q23 deletion, the pattern is one orange/green fusion signal

Full size image

Relationship of MYCN and 11q23 statuses with clinicobiological characteristics of patients with bone marrow metastatic neuroblastoma

We observed no association of MYCN amplification with age, sex, or primary tumor site (Table 2). However, the primary tumor site of all patients with MYCN amplification was in the abdomen. We found that MYCN amplification mainly occurred in those with high LDH and NSE levels (both P < 0.001), and more events were observed in patients with MYCN amplification than in those with normal MYCN (P = 0.004).

Table 2 Clinical characteristics of bone marrow metastatic neuroblastoma patients according to MYCN and 11q23 statuses
Full size table

Unlike the MYCN gene, 11q23 deletion was associated with age (P = 0.024). We found 38 (95.0%) cases of 11q23 deletion occurred in children older than 18 months. The median age at diagnosis was 41 months for patients with 11q23 deletion, but 24.5 months for those with MYCN amplification. Although 24 (60.0%) patients with 11q23 deletion had events occurred, no significant difference was observed between patients with 11q23 deletion and normal 11q23 (P = 0.118).

Impact of MYCN and 11q23 on prognosis of bone marrow metastatic neuroblastoma

The 3-year EFS and OS rates of the 101 patients were 45.9 ± 5.5% and 59.1 ± 5.2%, with a median follow-up time of 23 months (range, 4–44 months). The estimated 3-year EFS and OS rates of patients with MYCN amplification were obviously lower than those of patients with normal MYCN (8.3 ± 8.0% vs. 43.8 ± 8.5% and 10.4 ± 9.7% vs. 63.5% ± 5.7%, both P < 0.001) (Fig. 2a, b). The difference in survival was not significant between patients with and without 11q23 deletion (Fig. 2c, d). The median EFS of patients with and without MYCN amplification was 12.5 and 25 months, respectively. The median EFS of patients with and without 11q23 deletion was 20.5 and 26 months, respectively.

Fig. 2
figure 2

The prognostic significance of MYCN amplification and 11q23 deletion in patients with bone marrow metastatic neuroblastoma. a, b Event-free survival (EFS) and overall survival (OS) of the 101 patients stratified by MYCN gene. c, d EFS and OS of the 101 patients stratified by 11q23 status. e, f EFS and OS of the 89 patients without MYCN amplification stratified by 11q23 status. g, h EFS and OS of the 101 patients stratified by combined assessment of MYCN and 11q23 statuses

Full size image

To better understand the effect of 11q23 deletion on prognosis of bone marrow metastatic neuroblastoma, we performed further survival analysis of the 89 patients without MYCN amplification. We surprisingly found that 11q23 deletion was a powerful prognostic factor if the influence of the MYCN gene was excluded. The estimated 3-year EFS and OS rates of the 38 patients with 11q23 deletion were significantly lower than those of the 51 patients without 11q23 deletion (34.3 ± 9.5% vs. 53.4 ± 10.3%, P = 0.037; 42.9 ± 10.4% vs. 75.9 ± 6.1%, P = 0.048) (Fig. 2e, f).

We then investigated the prognostic value of the combination of MYCN and 11q23. The 101 patients were stratified into four groups according to the statuses of MYCN and 11q23. The four groups differed significantly in survival (both P < 0.001) (Fig. 2g, h). 51 patients without MYCN amplification and 11q23 deletion had the best prognosis (3-year EFS and OS rates were 53.4 ± 10.3% and 73.0 ± 6.2%). The 2 patients with both MYCN amplification and 11q23 deletion had the worst outcome (3-year EFS and OS rates were both 0). Multivariate analysis showed that only MYCN amplification was an independent prognostic factor of EFS for all patients (Table 3). 11q23 deletion was a prognostic factor in univariate analysis only for patients without MYCN amplification, but it was not an independent prognostic factor in the multivariate analysis.

Table 3 Univariate and multivariate analysis of event-free survival (EFS) of children with bone marrow metastatic neuroblastoma
Full size table

Discussion

In recent years, the pretreatment risk stratification for neuroblastoma patients has become increasingly necessary [26, 27]. Currently, the identified biological factors closely related to prognosis include age, histology, ploidy, MYCN gene and specific segmental chromosomal aberrations [28, 29]. In the present study, we evaluated the prognostic value of 11q23 deletion with or without MYCN amplification in bone marrow metastatic neuroblastoma. We found that 11q23 deletion was a prognostic marker only for patients without MYCN amplification and that 11q23 deletion was closely associated with age. Almost all cases of 11q23 deletion occurred in children older than 18 months. Multivariate analyses for EFS indicated that MYCN amplification was an independent prognostic factor, but 11q23 deletion was not. Combined assessment of MYCN and 11q23 deletion was much better than single-marker assessment in identifying the patients at a high risk of disease progression in bone marrow metastatic neuroblastoma.

In the present study, MYCN amplification was present in 11.9% of the patients, which was lower than those in our previous study [17] and other reports [21, 30]. 11q23 deletion was found in 39.6% of the patients, a rate close to those reported by other groups [11, 31]. Similar to another group [31], we here found that 11q23 deletion occurred commonly in patients with normal MYCN. In fact, the two markers were mutually exclusive; only 2 patients presented with both. We thought that MYCN amplification and 11q23 deletion were distinct genetic aberrations with different clinicobiological characteristics. MYCN amplification was associated with LDH and NSE levels, whereas 11q23 deletion was associated with age. There was a marked difference in age at diagnosis between the two groups, with a median age of 24.5 months in the MYCN amplification group and 41 months in the 11q23 deletion group.

We further confirmed that 11q23 deletion constituted a distinct group of patients with unfavorable neuroblastoma. In the present study, MYCN amplification was confirmed as an independent prognostic factor. We found no obvious differences in EFS and OS between the 11q23 deletion and the 11q23 normal groups. However, we found that 11q23 deletion was a powerful marker for poor prognosis in patients without MYCN amplification. Among the 89 patients without MYCN amplification, the 3-year EFS and OS rates were significantly lower in patients with 11q23 deletion than in those without the deletion (both P < 0.05).

It was reported that 11q23 deletion may be involved in the dysregulation of important genes, including tumor suppressor genes and oncogenes [32]. One candidate gene is forkhead box R1 (FOXR1), which is located in 11q23.3. 11q23.3 deletion results in either mixed-lineage leukemia (MLL) or platelet activating factor acetylhydrolase 1b catalytic subunit 2 (PAFAH1B2) fusing to the proximal side of FOXR1. The new fusion transcripts stimulate the overexpression of FOXR1. Functional analyses have shown that FOXR1 overexpression could promote cell proliferation and may be a tumor-driving event [33]. Another important tumor suppressor gene is h2a histone family member X (H2AFX), which resides in the 11q23.2-q23.3 region. H2AFX participates in DNA double-strand breaks (DSBs) through phosphorylation. The loss of H2AFX promotes tumorigenesis due to defective DSB repair, increased radiation sensitivity, and genomic instability [10]. It was reported that more chromosomal breaks were seen in neuroblastoma patients with 11q23 deletion, which may be explained by genomic instability related to the loss of H2AFX gene [34]. This hypothesis also implied that 11q23 deletion occurred early in tumorigenesis. However, in the present study, the median age at diagnosis was 41 months for patients with 11q23 deletion and 24.5 months for those with MYCN amplification. The definitive role of 11q23 deletion in tumorigenesis of neuroblastoma still needs to be elucidated.

The present study showed that a large number of patients with MYCN amplification or 11q23 deletion had unfavorable prognosis. The survival rate was higher in patients with 11q23 deletion than in those with MYCN amplification. However, OS has been reported to be similar for patients with the two aberrations after 8 years of follow-up [10]. We further evaluated the prognostic value of combined assessment of MYCN and 11q23 statuses and found that the combined assessment was much superior to single-marker assessment in recognizing the patients at a high risk of disease progression.

The present study investigated the outcomes of bone marrow metastatic neuroblastoma patients with 11q23 deletion and of the rare patients with both MYCN amplification and 11q23 deletion. However, there were some limitations of this study. First, it is unclear how the results would be used to affect treatment. Second, the cases of both MYCN amplification and 11q23 deletion were limited. In our future study, we will focus on how further subdivision of these patients would alter treatment.

Conclusions

In the present study, we found that for patients with bone marrow metastatic neuroblastoma, 11q23 deletion predicted poor prognosis only in those without MYCN amplification. Combined assessment of the two markers was superior to single-marker assessment in identifying patients at a high risk of disease progression.