Synergistic drug combination screening using a nanodroplet processing platform to enhance neuroblastoma treatment in TH‐MYCN transgenic mice

Abstract

Neuroblastoma is a highly aggressive pediatric cancer with a poor prognosis, particularly in high‐risk (HR) cases characterized by MYCN amplification. The severe side effects associated with high‐dose chemotherapy further complicate treatment. Despite significant advancements in drug screening, traditional platforms remain limited due to their requirement for large cell quantities and their low translational success from bench to clinic. These limitations hinder the application of personalized medicine screening for patients with neuroblastoma. To address these challenges, we developed a Bioinspired Nanodroplet Processing (BioNDP) platform. This innovative platform allows for the simultaneous screening of multiple drug combinations while reducing the required number of cells to just 100 and minimizing assay volumes to 200 nL per well. Using BioNDP, we screened chemotherapeutic combinations of cyclophosphamide, doxorubicin, and vincristine in both the SK‐N‐DZ neuroblastoma cell line and primary neuroblastoma cells derived from TH‐MYCN transgenic mice. Our findings revealed a specific drug combination that exhibited significant synergistic cytotoxicity in neuroblastoma cells. This combination completely eradicated tumors and significantly improved survival rates in TH‐MYCN mice, without notable side effects. This study highlights the potential of the BioNDP platform in bridging in vitro and in vivo results, offering a promising strategy for personalized medicine in the treatment of HR neuroblastoma, with reduced toxicity and enhanced therapeutic efficacy.

---

Translational Impact Statements

This study leverages a nanodroplet processing platform to advance treatment strategies for high-risk neuroblastoma. Using primary neuroblastoma cells derived from a transgenic mouse model, the platform identified a low-dose, synergistic drug combination that eradicates tumors while minimizing toxicity. By overcoming the limitations of traditional screening methods, this innovative approach bridges laboratory and clinical research, offering a path toward safer and more effective personalized therapies for aggressive pediatric cancers.

---

INTRODUCTION

Drug screening is a crucial step in identifying potential therapeutic compounds. Despite decades of advancements in drug screening platforms, conventional methods still encounter significant limitations. A major challenge in translational medicine is the low success rate in translating findings from bench to clinical application,^1–3 often due to the complexity of the tumor microenvironment, disparities between preclinical models and human physiology, and intra-patient heterogeneity.^4–6 These challenges underscore the need for innovative approaches. Personalized drug screening (PDS), which often requires patient-derived samples, is increasingly recognized as a key advancement to address traditional method limitations. However, conventional cell-based assays, which require large quantities of cells, restrict the scope of drug screenings when working with limited patient-derived samples. To address this issue, we developed the Bioinspired Nanodroplet Processing (BioNDP) platform. This novel platform significantly reduces the required number of cells and enables simultaneous screening of multiple drugs.^7–10 The BioNDP platform, which integrates an array chip with an automatic liquid dispenser, facilitates drug combination screening using only 100 cells and a reaction volume of 200 nL per well. It allows for the evaluation of multiple drug combinations and has demonstrated effective drug screening for prostate and breast cancers, with efficacy validated in both xenograft mouse models and zebrafish models.^7,11 This innovative platform not only reduces cell consumption but also enhances the capability for extensive drug combination screening.

Neuroblastoma, a malignancy derived from neural crest cells, accounts for 8%–10% of childhood cancers and is responsible for over 15% of pediatric cancer-related deaths.^12–14 Patients with neuroblastoma are classified into low-risk, intermediate-risk, or high-risk (HR) categories based on factors such as tumor stage defined by the International Neuroblastoma Risk Group Staging System, age at diagnosis, and genetic abnormalities.^15,16 MYCN amplification is particularly associated with HR neuroblastoma. The current therapeutic approach for HR neuroblastoma involves the aggressive administration of high-dose chemotherapeutics, including cyclophosphamide (CP) + doxorubicin (DOX) + vincristine (VCR) (the combination is abbreviated CDV) and cisplatin + etoposide.^15,17–19 However, this treatment strategy often results in severe side effects, and responses can vary significantly among patients. These challenges underscore the urgent need for personalized and more effective combination chemotherapeutic strategies for neuroblastoma.²⁰

The TH-MYCN mouse model is widely used to study the mechanisms and therapeutic approaches for neuroblastoma.²¹ This genetically engineered model overexpresses the human MYCN oncogene under the control of the tyrosine hydroxylase (TH) promoter to specifically express in neural crest cells.^22–24 This genetic modification closely mimics the MYCN amplification observed in human neuroblastoma, leading to spontaneous tumor development.^25,26 Furthermore, TH-MYCN mice possess an intact immune system, allowing for comprehensive observations of drug responses and immune interactions.^27,28 The TH-MYCN model accurately replicates both the genetic and pathological features of human neuroblastoma while preserving a functional immune system. Consequently, this animal model serves as a precise and reliable platform for preclinical testing in neuroblastoma research.^29,30

In this study, we sought to establish a PDS framework for neuroblastoma using the BioNDP platform. Our primary objective was to identify a drug combination that achieves superior therapeutic outcomes with reduced dosages compared to conventional treatments while minimizing adverse effects. To evaluate this approach, we employed TH-MYCN mice as a model for neuroblastoma. Primary neuroblastoma cells were isolated from tumors in TH-MYCN mice and exposed to various combinations of CP, DOX, and VCR—three standard chemotherapeutic agents used for HR patients with neuroblastoma. The BioNDP platform was used to determine the optimal drug combination, which was then validated for efficacy in TH-MYCN mice. This study effectively bridges the gap between in vitro and in vivo research, establishing a robust preclinical model for developing personalized therapeutic strategies in neuroblastoma.

MATERIALS AND METHODS

Cell culture

The human neuroblastoma cell line SK-N-DZ (ATCC) and the human embryonic kidney cell line HEK-293 (ATCC) were maintained in Dulbecco's Modified Eagle Medium (DMEM, HyClone), supplemented with 10% fetal bovine serum (FBS, Gibco) and 1% penicillin–streptomycin (GeneTeks Bioscience). Cells were cultured in a humidified incubator with 5% CO₂ at 37°C. Isolated primary neuroblastoma cells were maintained in a specialized culture medium comprising 50% neurobasal medium (Gibco), 40% DMEM/F12 (Gibco), 10% FBS, 1X B-27 minus vitamin A (Gibco), 1X N-2 supplement (Gibco), 0.01 μg/mL epidermal growth factor (Gold Biotechnology), 0.02 μg/mL basic fibroblast growth factor (Gold Biotechnology), and 25 μg/mL primocin (InvivoGen).

The 384-well plate cell viability assay

The diluted drug solutions were dispensed into the wells using the HP D300 digital dispenser, followed by dehydration under sterile conditions in a biosafety cabinet. Subsequently, 2000 SK-N-DZ cells were seeded into each well containing 25 μL of assay media. The cells were incubated for 24 h in a humidified incubator at 37°C with 5% CO₂. Following incubation, ATP levels in each well were measured using the CellTiter-Glo® luminescent cell viability assay (Promega), adhering to the manufacturer's protocol. Luminescence, indicative of cell viability, was quantified using the PHERAstar Detection System (BMG Labtech).

Drug combination screening on the BioNDP platform

The entire procedure is illustrated in Figure S1A and follows the methodology outlined in our previous study.⁷ The chip was sterilized using ultraviolet radiation for 30 min before use. Drug droplets of varying concentrations were then dispensed into the corresponding wells using a customized liquid handling dispenser (Versa 10 spotter, Aurora Instruments Ltd., Vancouver, CA). The dispensing operation exhibited remarkable stability, with a coefficient of variation (CV) as low as 0.8% (Figure S1B). Each droplet was precisely calibrated to a volume of 200 nL. After drug dispensing, the chip was incubated at room temperature for 10 min to evaporate the droplets, thereby converting the drug into powder form. A 200 nL droplet containing 100 cells was then dispensed into each well of the chip, which was subsequently sealed with a polydimethylsiloxane (PDMS)-glass gasket. After 24 h of incubation, cell viability was assessed using the CellTiter-Glo® luminescent cell viability assay, following the manufacturer's instructions (Promega). The luminescence intensities of live cells were captured using a UVP ChemStudio Plus (Analytik Jena) and analyzed with VisionWorks software (Analytik Jena).

Primary tumor cell isolation

Tumor samples were obtained from TH-MYCN transgenic mice. Fresh neuroblastoma tumor samples were washed with phosphate-buffered saline (PBS) containing 25 μg/mL primocin and then cut into 1–3 mm pieces. The tumor fragments were mixed with a primary neuroblastoma cell culture medium containing a blend of enzymes (Miltenyi Biotec) and transferred into a gentleMACS™ C Tube (Miltenyi Biotec). Tumor cell dissociation was carried out using a dissociator (RWD Life Science) per the equipment manual. Following dissociation, the cell suspension was filtered through a 70 μm strainer, and red blood cells were removed using red blood cell (RBC) lysis buffer (Gibco). Tumor cells were then purified through a two-step isolation process. In the first step, non-tumor cells were eliminated using a non-tumor cell depletion kit (Miltenyi Biotec), following the manufacturer's instructions. In the second step, the cells were incubated with Biotin-GD2 antibody (BioLegend) and isolated using a biotin targeting kit (Miltenyi Biotec), following the manufacturer's protocol to purify primary neuroblastoma cells expressing GD2.

Flow cytometry

Cells were collected and resuspended in 1× PBS containing 0.5% bovine serum albumin. The cells were then incubated with the fluorescein isothiocyanate (FITC)-conjugated GD2 antibody (BioLegend, 357313). Flow cytometric analysis was conducted using a fluorescence-activated cell sorting (FACS) Verse™ flow cytometer (BD Biosciences), and the data were analyzed with BD FACSuite software.

Animal experiment

This study adhered to the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The 129/SvJ-Tg(TH-MycN) mouse strain^22,31 was kindly provided by Prof. Akira Nakagawara (MD, PhD, Chiba Cancer Center, Chiba, Japan). The breeding of female TH-MYCN mice was conducted per the approved protocols of the Institutional Animal Care and Use Committee, College of Medicine, National Taiwan University. Hemizygous TH-MYCN mice were provided by the Transgenic Mouse Core Facility at the College of Medicine, National Taiwan University. After weaning, hemizygous TH-MYCN mice were monitored by ultrasound at the preaortic and adrenal areas to monitor spontaneous tumor growth. Sonograms were captured using the VisualSonics VEVO-2100 High-Frequency Ultrasound system (VisualSonics, Toronto, Canada). The screening strategy was optimized from Teitz et al.¹² In this study, animals were anesthetized with 1.5% isoflurane in O₂ at a flow rate of 2 L/min. Hair removal cream (Nair) was applied to the abdominal and thoracic areas. For ultrasound scanning, animals were placed in a supine position on the imaging stage, which was coated with ultrasound transmission gel, and the ultrasound transducer (RMV-706 at 40 MHz) was lowered stereotactically to the surface of the animal. Tumors, identified by irregular regions, were distinguished from normal tissue landmarks. During the imaging process, the maximal diameter of tumors was measured from axial and sagittal views, and tumor volume was calculated using the ellipsoid Equation (1)^32,33: 1 $\frac{1}{2}\times \text{width}\times \text{depth}\times \text{height}.$

This approach has been widely used in the literature for tumor volume measurement in TH-MYCN mouse models.³⁰

Treatment was initiated, and treatment response was evaluated once to twice a week with ultrasound imaging once the maximal diameter of the tumor achieved 0.5 cm or larger. These mice were randomly assigned to receive intraperitoneal injections of the following treatments: vehicle (0.9% saline) for 3 days, 12 mg/kg CP (Baxter Healthcare Corporation) for 2 days, 0.6 mg/kg DOX (Pfizer) for 3 days, 0.06 mg/kg VCR (Pfizer) for 3 days, or a combination of these three drugs (Figure S2). The treatment schedule was based on a commonly used regimen in recent clinical trials and the National Comprehensive Cancer Network clinical practice guidelines for HR neuroblastoma.^15,34 The study was conducted over 12 weeks, with mice humanely euthanized if the maximal diameter of tumors reached 1.5 cm during this period.

The tumor response percentage for each mouse was calculated using the following Equation (2): 2 $\text{Tumor response}\%=\frac{b_{\mathrm{x}}-a}{a}\times 100\%,$ where a represents the tumor volume measured on the day when the tumor's maximum diameter exceeds 0.5 cm, x is the number of days post-administration, and b is the tumor volume on day x.

The schedule for biochemical and hematological analyses in mice was aligned with the in vivo efficacy study described above. Whole blood and serum samples were collected from all mice 1 day after the final administration and analyzed by Bio-Cando Incorporation (Taoyuan, Taiwan).

Tumors and organs, including the liver, spleen, and kidneys, were collected from mice on the day of sacrifice upon reaching the experimental endpoint. The organ samples were sent to the National Taiwan University Animal Resource Center for fixation, slicing, and hematoxylin and eosin (H&E) staining, and the histological examination was performed by Dr. Chen from Chia-Yi Christian Hospital.

In vivo dosage calculation for animal experiments

The calculation process, which has been reported in our previous study, considers key factors including the Dose-Reduction Index (DRI), maximum tolerated dose (MTD), and administration frequency (n). DRI values were calculated based on the single-drug dose–response experiments of CP, DOX, and VCR using primary neuroblastoma cells on the BioNDP platform. The single-drug response data were presented in Figure S4. Drug combination studies were subsequently performed on the same platform. The drug concentration-effect data (cytotoxicity) were input into CompuSyn software,^35–37 which calculated the DRI values for the three drugs: CP (3.16), DOX (6.74), and VCR (139.50). The averaged DRI across the combination was 49.8, reflecting the extent of dosage reduction enabled by the synergy of the drug combination. The MTD values of murine for CP, DOX, and VCR were obtained from the literature.^38–40 Based on these studies, the following MTD values were used: CP: 300 mg/kg; DOX: 10 mg/kg; VCR: 1 mg/kg. The total number of drug administrations was based on the experimental design: CP: administered for 2 days over the treatment period; DOX: administered for 3 days over the treatment period; VCR: administered for 3 days over the treatment period.

The in vivo dosage for each drug was calculated using the following Equation (3): 3 $\text{In vivo dosage}=\left[\frac{\mathrm{MTD}}{\text{Averaged}\mathrm{DRI}}\right]\times n.$

Using the formula, the dosages were determined as follows: CP: 300/49.8 × 2 = 12 mg/kg; DOX: 10/49.8 × 3 = 0.6 mg/kg; VCR: 1/49.8 × 3 = 0.06 mg/kg.

Statistical analysis

One-way analysis of variance; This is a statistical tool for means comparison if different groups (ANOVA) was conducted to compare data across multiple groups and determine statistical significance. The half-maximal inhibitory concentration (IC₅₀) values were fitted and evaluated using a variable slope (four-parameter) logistic equation in GraphPad Prism 9.0.0. The CV percentage of droplet dispensing on the chip was calculated based on luminescence intensity using the following Equation (4). 4 $\mathrm{CV}\%=\frac{\text{standard deviation}}{{\text{Signal}}_{\mathrm{ave}}}\times 100\%.$

The percentage of live cells, derived from the detected luminescence intensity, was calculated using the following Equation (5): 5 $\text{Cell viability}\%=\frac{\gamma -\beta }{\alpha -\beta }\times 100\%.$ where α represents the luminescence signal from wells without any treatment, β represents the luminescence signal from wells without cells (serving as the background signal), and γ represents the luminescence signal from wells with the selected treatments.

Drug combination synergy was assessed using Combination Index (CI) values calculated with CompuSyn³⁵ software, classifying interactions as synergistic (CI <1), additive (0 ≤ CI ≤ 1), or antagonistic (CI >1).

RESULTS

Synergistic chemo-drug combination screening in SK-N-DZ neuroblastoma cell line through BioNDP platform

Previously, we successfully demonstrated the applicability of the BioNDP platform for drug screening in breast and prostate cancers.⁷ In this study, we expanded its application to neuroblastoma by evaluating the efficacy of DOX in the SK-N-DZ cell line, which is characterized by MYCN amplification. Cells were treated with DOX in 384-well plates and on BioNDP chips, followed by a 24-h incubation period. Subsequently, cell viability was assessed using the CellTiter-Glo assay, which measures ATP levels as an indicator of metabolically active cells. The IC₅₀ values were found to be comparable between the two platforms: 1.3 μg/mL for the 384-well plate and 2.3 μg/mL for the BioNDP chip, underscoring the reliability of BioNDP in drug assessment (Figure 1a). To determine the appropriate concentrations of CP and VCR for subsequent combination screenings, we evaluated drug responses in SK-N-DZ cells using the BioNDP platform. The IC₅₀ values for CP and VCR were determined to be 2.0 mg/mL and 2.1 μg/mL, respectively (Figure S3). Based on these efficacy tests, two concentrations of each drug were selected for screening potential synergistic combinations on the BioNDP platform. Notably, the combination of 1 mg/mL CP, 2 μg/mL DOX, and 2 μg/mL VCR resulted in a significant reduction in cell viability to 7.6%, indicating optimized cytotoxicity in SK-N-DZ cells (Figure 1b,c). In contrast, individual treatments exhibited lower cytotoxic effects (Figure 1d), highlighting the enhanced cytotoxic potential of the selected drug combination. To evaluate the synergistic effect of the drug combinations, we calculated the CI value, which quantitatively determines drug interactions.⁴¹ The CI value for the combination of 1 mg/mL CP, 2 μg/mL DOX, and 2 μg/mL VCR was 0.4 (Figure 1c), indicating a synergistic cytotoxic effect on SK-N-DZ cells. In summary, these initial results underscore the robustness and potential feasibility of the BioNDP platform in identifying synergistic drug combinations for neuroblastoma.

[IMAGE OMITTED. SEE PDF]

Identification for an optimal synergistic drug combination in primary neuroblastoma cells isolated from TH-MYCN transgenic mice

To further assess the feasibility of the personalized BioNDP neuroblastoma drug screening system, we utilized TH-MYCN mice as a preclinical model to replicate the conditions observed in patients with neuroblastoma. Initially, we established a two-step purification method for isolating primary neuroblastoma cells. This method consistently yielded purified cells exhibiting the morphological characteristics of adrenergic neuroblastoma cell types (Figure 2a), with high expression levels of GD2, a surface marker specific to neuroblastoma, as confirmed by FACS analysis. The percentage of GD2-positive cells in both isolated primary tumor cells (97.6%) and SK-N-DZ cells (98.4%) was significantly higher than in HEK-293 cells (0.1%) (Figure 2b). These findings indicate that this two-step selection method effectively isolates primary neuroblastoma cells with high purity.

[IMAGE OMITTED. SEE PDF]

Building on preliminary assays using SK-N-DZ cells as the drug target, we further explored effective doses for the combinational testing of CP, DOX, and VCR on primary neuroblastoma cells using the BioNDP platform. The IC₅₀ values were determined to be 3.38 mg/mL for CP, 4.55 μg/mL for DOX, and 13.27 μg/mL for VCR. Concentrations near the IC₅₀ of each drug were selected to assess combination effects, specifically 1, 2, and 4 mg/mL of CP; 1, 2, and 4 μg/mL of DOX; and 1, 2, and 8 μg/mL of VCR. The results demonstrated that the combination of 2 mg/mL CP, 4 μg/mL DOX, and 2 μg/mL VCR significantly reduced cell viability to 16.1%, showing higher cytotoxicity compared to single-drug treatments or other combinations (Figure 2c,d). Moreover, the CI value for this combination was 0.5, indicating a synergistic cytotoxic effect in primary neuroblastoma cells.

Evaluating the in vivo efficacy of a refined drug combination using TH-MYCN mice

To assess the in vivo therapeutic efficacy of the drug combination identified through the BioNDP platform, we administered the therapy to TH-MYCN mice. This administration protocol was adapted from the in vitro drug combination, following the algorithm developed in our previous study.⁷ Mice were systematically monitored using the VisualSonics VEVO-2100 High-Frequency Ultrasound system to detect tumor development, growth, or regression. Dosages of CP, DOX, and VCR for the animal experiments were determined based on the previously established algorithm⁷ and administered individually or in combination once tumors reached at least 0.5 cm in diameter (Figure 3a–e). The drug delivery procedure is illustrated in Figure S2. Following treatment, tumor sizes were monitored weekly using the ultrasound system. On average, tumor volumes in the vehicle-treated mice increased more than 15-fold within 2 weeks post-administration (Figure 3f). In contrast, tumors in mice receiving single-drug treatments continued to grow post-administration (Figure 3g–i), increasing to over sixfold their original size within 2 weeks post-administration (Figure 3k). Due to the rapid progression of tumors, several mice in the vehicle or single-drug treatment groups required early euthanasia in accordance with animal care guidelines, resulting in the premature termination of the experiments (Figure 3k). In contrast, mice treated with the drug combination exhibited complete tumor regression within 2 weeks (Figure 3j,k). However, we observed tumor relapse in two mice during the fourth week post-combination treatment, though re-administration of the same combination led to sustained tumor regression until the study concluded (Figure 3k). At the endpoint of the experiment, significant tumor growth was observed in mice treated with the vehicle or single drugs, whereas those treated with the selected combination showed no detectable tumors (Figure 3l). Furthermore, survival analysis revealed that this combination therapy significantly extended the lifespan of TH-MYCN mice compared to other treatments, in which mice either succumbed shortly after treatment or required euthanasia due to aggressive tumor growth (Figure 4a). To compare the dosages used in this study with clinically employed ones, we converted the dosages to human equivalent doses (HED) according to U.S. Food and Drug (FDA) guidelines.⁴² The conversion revealed that the HED of CP, DOX, and VCR in our study were approximately 1.72%, 7.35%, and 25% of the clinical doses, respectively (Figure 4b).

[IMAGE OMITTED. SEE PDF]

[IMAGE OMITTED. SEE PDF]

Assessment of side effects in TH-MYCN mice treated with the refined drug combination

Severe side effects are commonly associated with current treatment protocols for HR neuroblastoma, primarily due to the high doses of chemotherapeutic agents used in patients.⁴³ The drug combination identified by the BioNDP platform significantly reduced the required doses of CP, DOX, and VCR, potentially mitigating the incidence of side effects. To verify this potential reduction, we evaluated the side effects of various treatments in TH-MYCN mice, focusing on body weight changes and comprehensive blood analyses. Post-administration, no significant changes in body weight were observed across all treatment groups (Figure 5a). Additionally, biochemical and hematological analyses of blood samples from mice treated with either the vehicle or combination therapy revealed that markers such as alanine aminotransferase (ALT), aspartate aminotransferase (AST), creatinine (CREA), and blood urea nitrogen (BUN)^44,45 remained within normal physiological ranges, indicating no significant liver or kidney toxicity (Figure 5b, top panel). Moreover, the cell counts for various blood cell types in mice treated with the combination therapy were comparable to those in the vehicle control group, suggesting that the selected drug combination maintained normal blood homeostasis in the murine models (Figure 5b, bottom panel).^46,47 On the other hand, liver, kidney, and spleen samples were collected, fixed, and stained with H&E. After reviewing the histological features of major organs (e.g., liver, kidney, and spleen), we did not find evidence of hepatotoxicity or renal toxicity in single-agent therapies or combined therapy (Figure S6). These findings demonstrate that the drug combination identified by the BioNDP platform effectively inhibited tumor growth and induced tumor regression with minimal side effects.

[IMAGE OMITTED. SEE PDF]

Collectively, these results confirm the significant therapeutic efficacy of the selected drug combination, which successfully eradicated spontaneously developed neuroblastoma in TH-MYCN mice. The administration of the selected combination at relatively low dosages suggests the potential of the BioNDP platform to identify novel drug dosages that minimize side effects, as corroborated by the blood and histological examinations. Overall, the in vivo results validate the therapeutic efficacy of the drug combination identified via the BioNDP platform, underscoring its potential to effectively treat neuroblastoma and improve overall survival outcomes in TH-MYCN mice.

DISCUSSION

Traditional drug screening methods that utilize tumor cell lines in multi-well plates have been criticized for their lack of reliability in generating clinically beneficial results, primarily due to the genomic heterogeneity within patient populations. Moreover, a previous study suggested that ideal drug profiling should involve 2000–6000 tumor cell lines encompassing at least 20 different tissue origins⁴⁸—a requirement that significantly surpasses the capacity of existing screening platforms. Consequently, PDS, which employs patient-derived primary cells, has emerged as a pivotal advancement in therapeutic strategies. While primary cells such as blood cells and dermal fibroblasts are frequently applied in PDS platforms,⁴⁹ the acquisition of other primary cell types is often problematic, either due to their scarcity or limited availability, thereby posing significant challenges for compound screening. Conversely, existing PDS methods, including patient-derived tumor organoids⁵⁰ and patient-derived xenografts,⁵¹ have shown the capability to replicate the complex characteristics of a tumor microenvironment. However, these approaches are hindered by limitations such as low throughput, high development costs, and the absence of a fully functional immune system.^52,53 In this study, we utilized our BioNDP platform as a PDS strategy to screen synergistic chemotherapeutic combinations specifically in primary HR neuroblastoma. The BioNDP platform is optimized for low cell numbers usage and multi-drug screening, effectively addressing the limitations inherent in traditional screening platforms. We first validated the reliability of the BioNDP platform by comparing its performance to that of traditional 384-well plates using the SK-N-DZ cell line. The results demonstrated comparable efficacy in cytotoxicity between the two platforms (Figure 1a). Additionally, while a prior study demonstrated that a PDMS chip could handle 16,000 primary cancer cells per single-dose treatment,⁵⁴ our approach significantly reduced the required cell number to just 100 cells per well, representing almost a 0.5% reduction compared to the previous study. Furthermore, our platform facilitates the assessment of synergistic effects from drug combinations, which is critical in identifying effective treatment strategies for HR neuroblastoma. The BioNDP platform enables rapid drug screening using patient-derived samples, providing advantages such as lower costs and greater flexibility compared to traditional methods. These features collectively position the BioNDP platform as an ideal tool for initiating PDS, overcoming the existing limitations of traditional methods while significantly enhancing both efficiency and scalability. The platform's capacity for efficient and reliable PDS highlights its potential as a transformative tool in oncology.

HR neuroblastoma accounts for over 50% of newly diagnosed neuroblastoma cases. Intensive multi-agent chemotherapy is typically employed during the induction phase to reduce tumor burden in patients with HR neuroblastoma. Although higher doses of chemotherapy may offer the potential to cure HR neuroblastoma, most pediatric patients are unable to complete the treatment due to severe side effects. Consequently, there is an urgent need for more effective and tolerable medications. To address this challenge, a more advanced drug screening platform is essential. In this study, we screened a drug combination of CP, DOX, and VCR using the BioNDP platform, which has previously demonstrated efficacy in drug screening across multiple cancer types and has been validated by in vivo animal models.⁷ We successfully identified the combination of CP, DOX, and VCR (CDV) and determined its cytotoxic efficacy using a neuroblastoma cell line (Figure 1b–d) and primary cells isolated from TH-MYCN mice (Figure 2). Furthermore, this selected combination significantly reduced tumor volumes (Figure 3) and prolonged survival in the TH-MYCN mouse model (Figure 4a). These in vivo results demonstrated complete tumor eradication over 12 weeks, in stark contrast to the outcomes from single-drug or vehicle treatments, which resulted in continued tumor growth or mortality.

The Children's Oncology Group (COG) and the International Society of Pediatric Oncology Europe Neuroblastoma Group (SIOPEN) have developed various treatment regimens for HR neuroblastoma (Figure S5). In the current COG protocol, the CDV regimen includes CP at 70 mg/kg, DOX at approximately 0.68 mg/kg, and VCR at 0.022 mg/kg.⁴³ However, when these clinical dosages were converted to their equivalents for use in mice based on the formula from a previous study,⁴² the dosage for CP reached 861 mg/kg, significantly exceeding the MTD for mice. This discrepancy underscores the unsuitability of directly applying standard clinical protocols in murine studies. Additionally, the administration of CP, DOX, and VCR has been associated with liver toxicity, as evidenced by elevated levels of ALT and AST.⁴⁴ Moreover, treatment with CP at 50 mg/kg over 3 days led to a significant increase in CREA levels, indicating renal injury.⁵⁵ These findings raise concerns about the potential for severe liver and kidney damage in children treated with the current COG protocol. In contrast, the HED of the CDV combination identified through the BioNDP platform were 0.98 mg/kg for CP, 0.05 mg/kg for DOX, and 0.005 mg/kg for VCR (Figure 4b), substantially lower than those used in clinical practice.^43,56–63 Notably, mice treated with this combination maintained stable body weights, and comprehensive analyses of serum chemistry and blood counts revealed no significant hematological or non-hematological adverse effects throughout the study period (Figure 5a,b). Furthermore, the histological examinations demonstrated normal liver parenchyma and renal glomeruli and tubules in different groups of mice (Figure S6), along with our biochemical data on liver and renal functions. Collectively, pathological evaluations and biochemical tests have provided critical biosafety data for our mice model. These findings provide strong evidence to support the safety of the drug combination identified through the BioNDP platform. Previous studies have integrated multiple therapies, including radiopharmaceutical iodine-131 meta-iodobenzylguanidine,⁶² anaplastic lymphoma kinase (ALK) inhibitors,⁶⁴ autologous stem cell rescue,⁵⁷ and anti-GD2 monoclonal antibodies⁶⁵—into the intensive induction regimen of the CDV combination. Unfortunately, these regimens frequently resulted in Grades 3–4 hematological and non-hematological adverse effects during the clinical induction phase (Table 1). These observations highlight the critical requirements to reduce dosages during the induction phase while preserving therapeutic efficacy. Remarkably, the dosages administered in our study were 2- to 40-fold lower than those reported in earlier studies. Moreover, this treatment regimen did not produce significant side effects, as indicated by stable body weights, normal levels of serum markers and blood cell counts, and unremarkable pathological features of organs. In recent years, immunotherapies—particularly GD2-targeted therapies—have become increasingly central in the treatment of HR neuroblastoma. These therapies have been shown to enhance the effectiveness of chemotherapy when combined with other chemotherapeutic agents.^66,67 Additionally, PDMS-based chips have been employed to develop tumor-on-a-chip systems for screening combinations of chemotherapies and immunotherapies.⁶⁸ Our current study, combined with these advancements, suggests that the BioNDP platform could be instrumental in developing innovative screening methods for combining chemotherapy and immunotherapy in the treatment of HR neuroblastoma.

TABLE 1 Comparison of the human equivalent dose (HED) of the combination in this study with common dosages used in high-risk neuroblastoma patients.

Cyclophosphamide (mg/kg)	Doxorubicin (mg/kg)	Vincristine (mg/kg)	Regimen adjustment	Hematologic and non-hematologic adverse effects	Refs.
0.98 × 2 days (HED)	0.05 × 3 days (HED)	0.005 × 3 days (HED)	• Low dosages identified from the BioNDP platform	• No obviously observed	This study
70 × 2 days	2.0a × 3 days	0.067 × 3 days	• Reducing chemotherapy from seven to five cycles. • Immunotherapy with the anti-GD2 3F8 monoclonal antibody	• Severe neutropenia • Mucositis, hearing deficits	50
27.03a × 2 days	1.62 mg/kga × 3 days	0.04a × 3 days	• The first attempt by Pediatric Oncology Group (POG) to intensive chemotherapy in combination with autologous stem cell rescue (ASCR)	• Grades 3 or 4 toxicities due to induction	51
40.54a × 2 days	0.68a × 3 days	0.01a × 3 days	• Short topotecan-based induction regimen	• Febrile neutropenia • Mucositis and diarrhea	52
56.76a × 2 days	0.68a × 3 days	0.02a × 3 days	53
32.43a × 3 days	0.81a × 1 day	0.04a × 3 days	• Delayed intensification chemotherapy	• Grades 3 to 4 neutropenia, thrombocytopenia	54
70 × 2 days	0.68a × 3 days	0.022 mg/kg × 3 days	• Memorial Sloan Kettering Cancer Center (MSKCC)-N5 regimen	• Grades 3 to 4 hematologic • Infections, gastrointestinal side effects (stomatitis, nausea, vomiting, and diarrhea)	37
70 × 2 days	0.68a × 3 days	0.02a × 3 days	55
54.05a × 1 day	0.81a × 2 days	0.04a × 5 days	• Continual treatment of tumor cells with 131-I-MIBG radio-chemotherapy for 1 month	• Grade 4 neutropenia • Grade 1 vomiting and mucositis	56
32.4a × 1 day	NA	0.04a × 1 day	• Delayed local treatment	• Grade 4 hematological adverse effects • Grade 4 non-hematological adverse effects	57

More than half of the patients diagnosed with HR neuroblastoma experienced relapse despite undergoing comprehensive multimodal treatment, with the majority of recurrences occurring within the first 2 years post-diagnosis.⁶⁹ Previous studies have identified MYCN amplification as a key factor strongly associated with tumor relapse and aggressive growth.⁷⁰ In our current study, we observed a similar pattern: two mice experienced tumor relapse following the administration of the selected drug combination. Initially, tumor volumes completely regressed within 2 weeks of treatment (Figure 3k). However, relapse was observed in two out of the five treated mice 4 weeks later (Figure 3k, indicated by a star). We hypothesize that the selected combination successfully eradicated the tumor in three mice, but it is likely that “dormant” cancer cells persisted in the remaining two mice due to ongoing MYCN expression, allowing these cells to evade the treatment and cause the observed relapses (Figure 3k). Notably, when the same combination was administered again, the relapsed tumors regressed completely, with no further relapses observed until the end of the study. We further hypothesize that the low drug dosages used in our study prevented the development of drug resistance while simultaneously producing synergistic therapeutic effects. Although previous research suggests that CP may indirectly inhibit MYCN expression,⁷¹ the precise mechanisms underlying the synergistic interactions among CP, DOX, and VCR, and their effects on MYCN, remain unclear and warrant further investigation.

CONCLUSION

In our study, we utilized the BioNDP platform to screen drug combinations specifically targeting HR neuroblastoma, a cancer known for its poor prognosis and severe treatment-related side effects. This platform, which requires only a minimal number of cells for screening, demonstrated its effectiveness in the context of personalized medicine. Initially, we validated its efficacy using SK-N-DZ neuroblastoma cells. Subsequently, we identified a synergistic drug combination in primary tumor cells that significantly reduced tumor volumes in TH-MYCN transgenic mice. Remarkably, the combination treatment completely eradicated tumors in the mice over 12 weeks. Throughout the treatment, the mice maintained stable body weights, with serum markers and blood cell counts remaining within normal ranges and unremarkable pathological features of the organs. These findings underscore the BioNDP platform's potential to enhance neuroblastoma treatment outcomes by identifying effective, low-dose therapeutic combinations that minimize adverse effects.

AUTHOR CONTRIBUTIONS

Yen-Tzu Liao: Conceptualization; formal analysis; investigation; validation; visualization; writing – original draft; writing – review and editing. Zhi-Kai Yu: Investigation; validation. Yi-Xun Huang: Investigation; validation. Kuan-Hung Lin: Writing – review and editing. Ching-Te Kuo: Conceptualization; methodology; supervision; writing – review and editing. Tsai-Shan Yang: Methodology. Pei-Yi Wu: Methodology; resources; supervision; writing – review and editing. Chi-Tai Yeh: Resources. Yen-Lin Liu: Resources; methodology. Chien-Chin Chen: Conceptualization; methodology; validation; writing – review and editing. Chiung-Nien Chen: Conceptualization; funding acquisition; supervision. Wen-Ming Hsu: Conceptualization; funding acquisition; supervision; writing – review and editing. Hsinyu Lee: Conceptualization; funding acquisition; project administration; supervision; writing – review and editing.

ACKNOWLEDGMENTS

This work was supported by grant no. 112-2314-B-002-220-MY2 from the National Science and Technology Council (NSTC), Taiwan. We are grateful to the Transgenic Mouse Core Facility at the National Taiwan University College of Medicine Animal Experiment Center for their assistance in housing and breeding the transgenic mice utilized in this study. We also appreciate the Technology Commons in the College of Life Science at National Taiwan University for their technical support and Bio-Cando Corporation for their expert assistance in mouse blood collection and subsequent analyses of serum and whole blood. We also thank the National Taiwan University Animal Resource Center for their support in the processing and staining of animal tissue. All illustrative figures were created using .

CONFLICT OF INTEREST STATEMENT

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

DATA AVAILABILITY STATEMENT

Data sharing not applicable to this article as no datasets were generated or analysed during the current study.

References

Hay M, Thomas DW, Craighead JL, Economides C, Rosenthal J. Clinical development success rates for investigational drugs. Nat Biotechnol. 2014;32:40‐51.

Cook D, Brown D, Alexander R, et al. Lessons learned from the fate of AstraZeneca's drug pipeline: a five‐dimensional framework. Nat Rev Drug Discov. 2014;13:419‐431. doi:10.1038/nrd4309

DiMasi JA, Feldman L, Seckler A, Wilson A. Trends in risks associated with new drug development: success rates for investigational drugs. Clin Pharmacol Ther. 2010;87:272‐277. doi:10.1038/clpt.2009.295

Goetz LH, Schork NJ. Personalized medicine: motivation, challenges, and progress. Fertil Steril. 2018;109:952‐963. doi:10.1016/j.fertnstert.2018.05.006

Ho D, Quake SR, McCabe ERB, et al. Enabling technologies for personalized and precision medicine. Trends Biotechnol. 2020;38:497‐518. doi:10.1016/j.tibtech.2019.12.021

Almendro V, Marusyk A, Polyak K. Cellular heterogeneity and molecular evolution in cancer. Annu Rev Pathol Mech Dis. 2013;8:277‐302.

Kuo CT, Wang JY, Lu SR, et al. A nanodroplet cell processing platform facilitating drug synergy evaluations for anti‐cancer treatments. Sci Rep. 2019;9:10120. doi:10.1038/s41598‐019‐46502‐3

Kuo C‐T, Lee H, Lee S‐C. Evidence of “wired” drug‐cell communication through micro‐barrier well‐array devices. AIP Adv. 2019;9(9):095025. doi:10.1063/1.5115170

Kuo CT, Chen CC. Biomimetic wax interfaces facilitating rehealable polymer composites. Polymers. 2021;13(18):3052. doi:10.3390/polym13183052

Kuo CT, Wang JY, Wo AM, Chen BPC, Lee H. ParaStamp and its applications to cell patterning, drug synergy screening, and rewritable devices for droplet storage. Adv Biosyst. 2017;1(5):1700048. doi:10.1002/adbi.201700048

Kuo CT, Lai YS, Lu SR, Lee H, Chang HH. Microcrater‐arrayed chemiluminescence cell chip to boost anti‐cancer drug administration in zebrafish tumor xenograft model. Biology. 2021;11(1):4. doi:10.3390/biology11010004

Teitz T, Stanke JJ, Federico S, et al. Preclinical models for neuroblastoma: establishing a baseline for treatment. PLoS One. 2011;6:e19133. doi:10.1371/journal.pone.0019133

Johnsen JI, Dyberg C, Wickstrom M. Neuroblastoma‐a neural crest derived embryonal malignancy. Front Mol Neurosci. 2019;12:9. doi:10.3389/fnmol.2019.00009

Nakagawara A, Li Y, Izumi H, Muramori K, Inada H, Nishi M. Neuroblastoma. Jpn J Clin Oncol. 2018;48:214‐241. doi:10.1093/jjco/hyx176

National Comprehensive Cancer Network. NCCN clinical practice guidelines in oncology (NCCN Guidelines®) neuroblastoma. 2024 https://www.nccn.org/guidelines/guidelines-detail?category=1&id=1520. Accessed 15 May 2024.

Matthay KK, Maris JM, Schleiermacher G, et al. Neuroblastoma. Nat Rev Dis Primers. 2016;2:16078. doi:10.1038/nrdp.2016.78

Dalianis T, Lukoseviciute M, Holzhauser S, Kostopoulou ON. New approaches towards targeted therapy for childhood neuroblastoma. Anticancer Res. 2023;43:3829‐3839. doi:10.21873/anticanres.16570

Zhou X, Wang X, Li N, Guo Y, Yang X, Lei Y. Therapy resistance in neuroblastoma: mechanisms and reversal strategies. Front Pharmacol. 2023;14:1114295. doi:10.3389/fphar.2023.1114295

DuBois SG, Macy ME, Henderson TO. High‐risk and relapsed neuroblastoma: toward more cures and better outcomes. Am Soc Clin Oncol Educ Book. 2022;42:1‐13. doi:10.1200/edbk_349783

Wu PY, Chuang PY, Chang GD, et al. Novel endogenous ligands of aryl hydrocarbon receptor mediate neural development and differentiation of neuroblastoma. ACS Chem Nerosci. 2019;10:4031‐4042. doi:10.1021/acschemneuro.9b00273

Chen L, Esfandiari A, Reaves W, et al. Characterisation of the p53 pathway in cell lines established from TH‐MYCN transgenic mouse tumours. Int J Oncol. 2018;52:967‐977. doi:10.3892/ijo.2018.4261

Chesler L, Weiss WA. Genetically engineered murine models – contribution to our understanding of the genetics, molecular pathology and therapeutic targeting of neuroblastoma. Semin Cancer Biol. 2011;21:245‐255. doi:10.1016/j.semcancer.2011.09.011

Kiyonari S, Kadomatsu K. Neuroblastoma models for insights into tumorigenesis and new therapies. Expert Opin Drug Discovery. 2015;10:53‐62. doi:10.1517/17460441.2015.974544

Weiss WA, Aldape K, Mohapatra G, Feuerstein BG, Bishop JM. Targeted expression of MYCN causes neuroblastoma in transgenic mice. EMBO J. 1997;16(11):2985‐2995.

Jamin Y, Tucker ER, Poon E, et al. Evaluation of clinically translatable MR imaging biomarkers of therapeutic response in the TH‐MYCN transgenic mouse model of neuroblastoma. Radiology. 2013;266(1):130‐140. doi:10.1148/radiol.12120128

Yogev O, Almeida GS, Barker KT, et al. In vivo modeling of chemoresistant neuroblastoma provides new insights into chemorefractory disease and metastasis. Cancer Res. 2019;79:5382‐5393. doi:10.1158/0008‐5472.CAN‐18‐2759

Kroesen M, Nierkens S, Ansems M, et al. A transplantable TH‐MYCN transgenic tumor model in C57Bl/6 mice for preclinical immunological studies in neuroblastoma. Int J Cancer. 2014;134:1335‐1345. doi:10.1002/ijc.28463

Webb ER, Lanati S, Wareham C, et al. Immune characterization of pre‐clinical murine models of neuroblastoma. Sci Rep. 2020;10:16695. doi:10.1038/s41598‐020‐73695‐9

Lu J, Guan S, Zhao Y, et al. The second‐generation ALK inhibitor alectinib effectively induces apoptosis in human neuroblastoma cells and inhibits tumor growth in a TH‐MYCN transgenic neuroblastoma mouse model. Cancer Lett. 2017;400:61‐68.

Rasmuson A, Segerström L, Nethander M, et al. Tumor development, growth characteristics and spectrum of genetic aberrations in the TH‐MYCN mouse model of neuroblastoma. PLoS One. 2012;7:e51297.

Quarta C, Cantelli E, Nanni C, et al. Molecular imaging of neuroblastoma progression in TH‐MYCN transgenic mice. Mol Imaging Biol. 2013;15(2):194‐202. doi:10.1007/s11307‐012‐0576‐9

Tomayko MM, Reynolds CP. Determination of subcutaneous tumor size in athymic (nude) mice. Cancer Chemother Pharmacol. 1989;24:148‐154. doi:10.1007/bf00300234

Feldman JP, Goldwasser R, Mark S, Schwartz J, Orion I. A mathematical model for tumor volume evaluation using two‐dimensions. J Appl Quant Methods. 2009;4:455‐462.

Kreissman SG, Seeger RC, Matthay KK, et al. Purged versus non‐purged peripheral blood stem‐cell transplantation for high‐risk neuroblastoma (COG A3973): a randomised phase 3 trial. Lancet Oncol. 2013;14:999‐1008. doi:10.1016/S1470‐2045(13)70309‐7

Chou T‐C. The combination index (CI < 1) as the definition of synergism and of synergy claims. Synergy. 2018;7:49‐50. doi:10.1016/j.synres.2018.04.001

Tao L, Zhou Y, Pan X, et al. Repression of LSD1 potentiates homologous recombination‐proficient ovarian cancer to PARP inhibitors through down‐regulation of BRCA1/2 and RAD51. Nat Commun. 2023;14:7430. doi:10.1038/s41467‐023‐42850‐x

Bryant KL, Stalnecker CA, Zeitouni D, et al. Author correction: combination of ERK and autophagy inhibition as a treatment approach for pancreatic cancer. Nat Med. 2020;26(6):628‐640. doi:10.1038/s41591‐019‐0368‐8

Aston WJ, Hope DE, Nowak AK, Robinson BW, Lake RA, Lesterhuis WJ. A systematic investigation of the maximum tolerated dose of cytotoxic chemotherapy with and without supportive care in mice. BMC Cancer. 2017;17:684. doi:10.1186/s12885‐017‐3677‐7

Harrison SD Jr. An investigation of the mouse as a model for vincristine toxicity. Cancer Chemother Pharmacol. 1983;11:62‐65. doi:10.1007/bf00257421

De Witt M, Gamble A, Hanson D, et al. Repurposing mebendazole as a replacement for vincristine for the treatment of brain tumors. Mol Med. 2017;23:50‐56. doi:10.2119/molmed.2017.00011

Zhang N, Fu JN, Chou TC. Synergistic combination of microtubule targeting anticancer fludelone with cytoprotective panaxytriol derived from panax ginseng against MX‐1 cells in vitro: experimental design and data analysis using the combination index method. Am J Cancer Res. 2016;6:97‐104.

Nair AB, Jacob S. A simple practice guide for dose conversion between animals and human. J Basic Clin Pharmacol. 2016;7:27‐31. doi:10.4103/0976‐0105.177703

Garaventa A, Poetschger U, Valteau‐Couanet D, et al. Randomized trial of two induction therapy regimens for high‐risk neuroblastoma: HR‐NBL1.5 International Society of Pediatric Oncology European Neuroblastoma Group Study. J Clin Oncol. 2021;39(23):2552‐2563. doi:10.1200/jco.20.03144

David S, Hamilton JP. Drug‐induced liver injury. US Gastroenterol Hepatol Rev. 2010;6:73‐80.

Sujino Y, Nakano S, Tanno J, et al. Clinical implications of the blood urea nitrogen/creatinine ratio in heart failure and their association with haemoconcentration. ESC Heart Fail. 2019;6:1274‐1282. doi:10.1002/ehf2.12531

UCLA Division of Laboratory Animal Medicine. CBC and differential reference range. 2005. https://labs.dgsom.ucla.edu/dlam/files/view/docs/diagnostic‐lab‐services/private/hematology_reference_range.pdf. Accessed 8 Mar. 2024.

UCLA Division of Laboratory Animal Medicine. Serum chemistry reference range – mouse. 2013. https://labs.dgsom.ucla.edu/dlam/files/view/docs/diagnostic‐lab‐services/private/serum_chemistry_reference_ranges_mice.pdf. Accessed 8 Mar. 2024.

Sharma SV, Haber DA, Settleman J. Cell line‐based platforms to evaluate the therapeutic efficacy of candidate anticancer agents. Nat Rev Cancer. 2010;10:241‐253. doi:10.1038/nrc2820

Gorshkov K, Chen CZ, Marshall RE, et al. Advancing precision medicine with personalized drug screening. Drug Discov Today. 2019;24:272‐278.

Hicks WH, Bird CE, Gattie LC, et al. Creation and development of patient‐derived organoids for therapeutic screening in solid cancer. Curr Stem Cell Rep. 2022;8:107‐117.

Abdolahi S, Ghazvinian Z, Muhammadnejad S, Saleh M, Asadzadeh Aghdaei H, Baghaei K. Patient‐derived xenograft (PDX) models, applications and challenges in cancer research. J Transl Med. 2022;20:206. doi:10.1186/s12967‐022‐03405‐8

Jin J, Yoshimura K, Sewastjanow‐Silva M, Song S, Ajani JA. Challenges and prospects of patient‐derived xenografts for cancer research. Cancers. 2023;15(17):4352. doi:10.3390/cancers15174352

Gao J, Lan J, Liao H, et al. Promising preclinical patient‐derived organoid (PDO) and xenograft (PDX) models in upper gastrointestinal cancers: progress and challenges. BMC Cancer. 2023;23:1205. doi:10.1186/s12885‐023‐11434‐9

Wong AH‐H, Li H, Jia Y, et al. Drug screening of cancer cell lines and human primary tumors using droplet microfluidics. Sci Rep. 2017;7(1):9109. doi:10.1038/s41598‐017‐08831‐z

Alabi QK, Akomolafe RO, Omole JG, et al. Polyphenol‐rich extract of Ocimum gratissimum leaves prevented toxic effects of cyclophosphamide on the kidney function of Wistar rats. BMC Complement Med Ther. 2021;21:274. doi:10.1186/s12906‐021‐03447‐3

Kushner BH, Kramer K, LaQuaglia MP, Modak S, Yataghene K, Cheung NK. Reduction from seven to five cycles of intensive induction chemotherapy in children with high‐risk neuroblastoma. J Clin Oncol. 2004;22:4888‐4892. doi:10.1200/jco.2004.02.101

Zage PE, Kletzel M, Murray K, et al. Outcomes of the POG 9340/9341/9342 trials for children with high‐risk neuroblastoma: a report from the Children's Oncology Group. Pediatr Blood Cancer. 2008;51:747‐753. doi:10.1002/pbc.21713

De Loris MA, Castellano A, Ilari I, et al. Short topotecan‐based induction regimen in newly diagnosed high‐risk neuroblastoma. Eur J Cancer. 2011;47(4):572‐578. doi:10.1016/j.ejca.2010.10.023

Rujkijyanont P, Photia A, Traivaree C, et al. Clinical outcomes and prognostic factors to predict treatment response in high risk neuroblastoma patients receiving topotecan and cyclophosphamide containing induction regimen: a prospective multicenter study. BMC Cancer. 2019;19:961. doi:10.1186/s12885‐019‐6186‐z

Zhang YT, Chang J, Xu HM, Li YN, Zhong XD, Liu ZL. Treatment of neuroblastoma with a novel delayed intensification chemotherapy. Indian J Pediatr. 2019;86:126‐131. doi:10.1007/s12098‐018‐2737‐6

Lee AC, Chui CH, Kwok R, Lee KS, Fong CM, Wong WH. Treatment and outcomes of high‐risk neuroblastoma in Southeast Asia: a single‐institution experience and review of the literature. Singapore Med J. 2023;64:319‐325. doi:10.11622/smedj.2021164

Mastrangelo S, Attinà G, Zagaria L, Romano A, Ruggiero A. Induction regimen in high‐risk neuroblastoma: a pilot study of highly effective continuous exposure of tumor cells to radio‐chemotherapy sequence for 1 month. The critical role of iodine‐131‐metaiodobenzylguanidine. Cancers Basel. 2022;14(20):5170. doi:10.3390/cancers14205170

Yoneda A, Shichino H, Hishiki T, et al. A nationwide phase II study of delayed local treatment for children with high‐risk neuroblastoma: the Japan Children's Cancer Group Neuroblastoma Committee Trial JN‐H‐11. Pediatr Blood Cancer. 2024;71(6):e30976. doi:10.1002/pbc.30976

Tucker ER, Jiménez I, Chen L, et al. Combination therapies targeting ALK‐aberrant neuroblastoma in preclinical models. Clin Cancer Res. 2023;29:1317‐1331. doi:10.1158/1078‐0432.Ccr‐22‐2274

Mora J, Cruz O, Lavarino C, et al. Results of induction chemotherapy in children older than 18 months with stage‐4 neuroblastoma treated with an adaptive‐to‐response modified N7 protocol (mN7). Clin Transl Oncol. 2015;17:521‐529. doi:10.1007/s12094‐014‐1273‐8

Raiser P, Schleiermacher G, Gambart M, et al. Chemo‐immunotherapy with dinutuximab beta in patients with relapsed/progressive high‐risk neuroblastoma: does chemotherapy backbone matter? Eur J Cancer. 2024;202:114001. doi:10.1016/j.ejca.2024.114001

Mora J, Modak S, Kinsey J, Ragsdale CE, Lazarus HM. GM‐CSF, G‐CSF or no cytokine therapy with anti‐GD2 immunotherapy for high‐risk neuroblastoma. Int J Cancer. 2024;154:1340‐1364. doi:10.1002/ijc.34815

Mu P, Zhou S, Lv T, et al. Newly developed 3D in vitro models to study tumor‐immune interaction. J Exp Clin Cancer Res. 2023;42:81. doi:10.1186/s13046‐023‐02653‐w

Basta NO, Halliday GC, Makin G, et al. Factors associated with recurrence and survival length following relapse in patients with neuroblastoma. Br J Cancer. 2016;115:1048‐1057. doi:10.1038/bjc.2016.302

Herd F, Basta NO, McNally RJQ, Tweddle DA. A systematic review of re‐induction chemotherapy for children with relapsed high‐risk neuroblastoma. Eur J Cancer. 2019;111:50‐58. doi:10.1016/j.ejca.2018.12.032

Álvarez‐León W, Mendieta I, Delgado‐González E, Anguiano B, Aceves C. Molecular iodine/cyclophosphamide synergism on chemoresistant neuroblastoma models. Int J Mol Sci. 2021;22(16):8936. doi:10.3390/ijms22168936