Risk Model for Metastasis Detection of Neuroblastoma (NB)

Bone Marrow Cytology-based Artificial Intelligence Model for Detection and Prognosis of Neuroblastoma

Neuroblastoma (NB) is the most common extracranial solid tumor in children, accounting for about 15% of tumor-related mortality. NB patients in high-risk group are prone to bone marrow and/or bone metastases with low five-year overall survival rate. The artificial intelligence (AI) and deep learning technologies have potential to identifying morphological characteristics of bone marrow cytology in clinical practice. In this study, the investigators construct and evaluate the bone marrow cytology-based AI model for detection and prognosis of NB. The main questions of the study as follows:

The question 1: Dose bone marrow cytology-based AI model work for prediction of bone marrow metastasis in NB? The question 2: Dose bone marrow cytology-based AI model work for prediction of bone metastasis in NB? The question 3: Dose bone marrow cytology-based AI model have potential to assist doctors in making individualized predictions of survival outcome? The investigators will retrospectively obtain the participants with NB between January 2019 and June 2024. The follow-up date ended on June 30, 2024.

The internal cohort including participants from Xinhua Hospital, Shanghai Jiao Tong University School of Medicine. The independent external cohorts including participants form Children's Hospital, Zhejiang University School of Medicine and Shenzhen Children's Hospital.

The investigators collect the clinical data of enrolled participants at the time of the patients' initial admission to the hospital, prior to receiving treatment. The clinical information including age, gender, primary tumor location, tumor grade, bone marrow metastasis state, bone metastasis state, genetic aberrations (MYCN amplification, Chromosome 1p deletion, Chromosome 11q deletion) and lab variables (peripheral blood cell count, bone marrow cytology indicators, the serum concentration of lactate dehydrogenase, neuron specific enolase).

This study is a non-interventional observational study, there is no risk to the participants and investigators. Participants get these benefits:

1. Early Detection: The model helps in early risk identification and personalize treatment.
2. Convenience: Because the model relies on general lab tests, it is easy to carry out can reduce invasive diagnostic procedures.
3. Cost-Effective: Using existing clinical data from routine tests can make the prediction process more cost-effective.
4. Data-Driven Decisions: The AI model improve diagnostic efficiency and support the medical decision.

Study Overview

• Status: Enrolling by invitation
• Conditions: Bone Metastases; Prognosis; Neuroblastoma (NB); Bone Marrow Metastasis
• Intervention / Treatment: Other: risk model in diagnosis and prognosis

Detailed Description

1. Dataset processing. In this study, the model including training set, validation set and test sets. During the model construction, the investigators process the training in internal cohort by randomly assign the enrolled participants to the training set and validation set in a 8:2 ratio. The investigators use the enrolled participants of external cohorts as test sets to evaluate the model performance independently, and then select the best model for future use.

   The Ethics Committee of Xinhua Hospital Affiliated to Shanghai Jiaotong University School of Medicine apaproved this study (XHEC-C-2024-023-2).

2. image acquisition. In the AI model, the investigators use bone marrow smears of enrolled participants for cytological evaluation and image collection. During cell classification and metastasis detection, the experienced pathologists complete the bone marrow smear analysis, metastatic NB clusters in bone marrow usually exhibited aggregated round atypical cells with high nucleus/cytoplasmic ratio. The investigators scan stained bone marrow smear at 40 × magnification for digital whole slide imaging (WSI). The investigators segment WSIs into smaller patches as 512 × 512 pixels tiles and apply the Vahadane method to normalize the color of small tiles.

3. Deep learning training. In the feature extraction of bone marrow cytological image, the process including two tiers of predictions: patch-level and WSI-level predictions.

   For patch-level predictions, the investigators carry out label predictions and their respective probabilities for all patches. The investigators apply model in the deep learning process as follows: recognized neural network (CNN)-resnet50 and Vision Transformer (ViT). The parameter configurations in the model as follows: optimizer-SGD, loss function-softmax cross-entropy, with a batch size of 64.

   For WSI-level predictions, the investigators use a multi-instance learning (MIL) algorithm to aggregate dispersed patch-level features to WSI-level features. During MIL for WSI fusion, the investigators perform WSI-level predictions with Patch Likelihood Histogram (PLH) pipeline and Bag of Words (BoW) pipeline in combination. Subsequently, the investigators get the WSI-level prediction as final representations of participant for subsequent analytical operations.

4. Signature building. In the procession of feature selection, the investigators use LASSO (Least Absolute Shrinkage and Selection Operator) feature screening to determine the final WSI-level features of the bone marrow cytology. These selected features were then subjected to machine learning methods to develop AI model. the investigators apply several machine learning algorithms to predict metastasis state of bone marrow and/or bone in participants, such as support vector machines (SVM), Logistic regression (LR), tree-based models, such as random forests and extremely randomized trees (ExtraTrees), extreme Gradient Boosting (XGBoost), and light gradient boosting machine (LightGBM), as well as multilayer perceptron (MLP) to develop our models.

   In the prognosis model, the investigators e use Cox models to construct the survival model with bone marrow cytological signature and clinical characters.

5. Model evaluation and statistical analysis. The investigators compare the clinical characteristics of participant with independent sample t-test for continuous variables and the χ² test for discrete variables in SPSS version 22.0 (SPSS, Inc., Chicago, IL, USA). P ≤ 0.05 was considered statistically significant.

For the diagnostic model, the investigators use both micro and macro area under the curve (AUC) metrics to evaluate the model in terms of sensitivity, specificity, accuracy, positive predictive value and negative predictive value at different classification thresholds. The model's performance and effectiveness were evaluated on separate test cohort. The study employs custom Python code written in Python v.3.7.12 to evaluate the model performance.

For the prognostic model, the investigators use AUC as the performance metric and calculating sensitivity and specificity. The model with the best performance on the test set was selected as the optimal model. Survival curves were constructed according to the Kaplan-Meier method.

• Study Type: Observational
• Enrollment (Estimated): 500

Contacts and Locations

Study Locations

• China
  • Shenzhen, China, 518038
    • Shenzhen Children's Hospital
  • Shanghai
    • Shanghai, Shanghai, China, 200092
      • Xinhua Hospital Affiliated to Shanghai Jiaotong University School of Medicine
  • Zhejiang
    • Hangzhou, Zhejiang, China
      • The Children's Hospital, Zhejiang University School of Medicine

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Child; Adult; Older Adult
• Accepts Healthy Volunteers: No

• Sampling Method: Non-Probability Sample

Study Population

This study included participants who were newly diagnosed with NB between January 2019 and June 2024. The follow-up date ended on June 30, 2024.

The internal cohort including participants from Xinhua Hospital, Shanghai Jiao Tong University School of Medicine. The independent external cohorts including participants form Children's Hospital, Zhejiang University School of Medicine and Shenzhen Children's Hospital. All the participants have performed bone marrow smear analysis as routine examination. The bone marrow smear stained with Wright-Giemsa was made according to standard protocols.The diagnosis completed by experienced pathologist and correlated with clinical and/or radiological findings.

Description

Inclusion Criteria:

1. The participant newly diagnosed with NB according to the International Neuroblastoma Risk Group Staging System (INRGSS). The diagnosis completed by experienced pathologist and correlated with clinical and/or radiological findings.
2. The participant diagnosed with NB at other hospitals who have not received chemotherapy or radiotherapy.
3. The participant with NB has performed bone marrow smear analysis as routine examination. The bone marrow smear stained with Wright-Giemsa was made according to standard protocols.

Exclusion Criteria:

1. The participant with concurrent diagnosis of other malignancies.
2. The participant with NB who has previously received chemotherapy and/or radiotherapy.
3. The participant with incomplete clinical data, the metastasis state of bone marrow and/or bone is unclear.
4. The participant was excluded due to non-representative specimens, such as unclear or faded Wright-Giemsa staining of bone marrow smear.

Study Plan

How is the study designed?

Number of groups / cohorts

2

Cohorts and Interventions

Group / Cohort	Intervention / Treatment
Neuroblastoma With Bone Marrow Metastasis Group; For the diagnosis of neuroblastoma with bone marrow metastasis, the medical practices including as follows: bone marrow biopsy, bone marrow cytology of aspiration smear, flow cytometry and positron emission tomography-computed tomography(PET-CT). Bone marrow biopsy or smear analysis may reveal characteristic NB cells. Flow cytometry may detect NB cells with phenotype of cluster of differentiation antigen 45(CD45)-/cluster of differentiation antigen 56(CD56)+/cluster of differentiation antigen 81(CD81)+/GD2 ganglioside (GD2)+. PET/CT imaging reveal the metastatic NB cells in term of metabolic activity and spatial distribution of metastatic involvement. A positive result from any of these methods is sufficient for diagnosed as NB with bone marrow metastasis.	Other: risk model in diagnosis and prognosis; In this study, we construct and evaluate the bone marrow cytology-based AI model for detection and prognosis of NB.; For the diagnostic model, we use AUC metrics to evaluate the model in terms of sensitivity, specificity, accuracy, positive predictive value and negative predictive value at different classification thresholds.; For the prognostic model, we use AUC as the performance metric and calculating sensitivity and specificity. Survival curves were constructed according to the Kaplan-Meier method.
Neuroblastoma Without Bone Marrow Metastasis Group; For the diagnosis of bone marrow metastasis in the enrolled participants, if there is no positive result from any of these tests as follows: bone marrow biopsy, bone marrow cytology of smear, flow cytometry or PET/CT, the participant is classified into the Neuroblastoma Without Bone Marrow Metastasis Group.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Neuroblastoma with bone marrow metastasis	the medical practices in diagnosis of bone marrow metastasis including as follows: bone marrow biopsy, bone marrow cytology of aspiration smear, flow cytometry and PET/CT. Bone marrow biopsy or smear analysis may reveal characteristic NB cells. Flow cytometry may detect NB cells with phenotype of CD45-/CD56+/CD81+/GD2+. PET/CT imaging reveal the metastatic NB cells in term of metabolic activity and spatial distribution of metastatic involvement. A positive result from any of these methods is sufficient for diagnosed as NB with bone marrow metastasis.	The period from the initial diagnosis of neuroblastoma to the initiation of chemotherapy or radiotherapy, up to 3 months.
Neuroblastoma with bone metastasis	We diagnosed NB with bone metastasis if bone destruction or discontinuity of the bone cortex in radiology test including CT, PET/CT, MRI.	The period from the initial diagnosis of neuroblastoma to the initiation of chemotherapy or radiotherapy, up to 3 months.

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Overall survival time	Overall survival (OS) was defined as time from diagnosis of neuroblastoma to death from any cause.	through study completion, up to 60 months.

Collaborators and Investigators

• Sponsor: Xinhua Hospital, Shanghai Jiao Tong University School of Medicine
• Collaborators: The Children's Hospital of Zhejiang University School of Medicine; Shenzhen Children's Hospital
• Investigators: Principal Investigator: juan ma, Doctor, Xin Hua Hospital, Shanghai Jiao Tong University School of Medicine

Study record dates

Study Major Dates

• Study Start (Actual): December 1, 2024
• Primary Completion (Estimated): July 1, 2025
• Study Completion (Estimated): December 1, 2026

Study Registration Dates

• First Submitted: November 19, 2024
• First Submitted That Met QC Criteria: November 21, 2024
• First Posted (Actual): November 25, 2024

Study Record Updates

• Last Update Posted (Actual): May 15, 2025
• Last Update Submitted That Met QC Criteria: May 12, 2025
• Last Verified: December 1, 2024

More Information

Terms related to this study

• Keywords: Prognosis; Neuroblastoma; Bone Metastasis; Bone Marrow Metastasis; Machine Learning Algorithms; Risk Prediction Models; artificial intelligence mode
• Additional Relevant MeSH Terms: Pathologic Processes; Neoplasms; Neoplasms by Histologic Type; Neoplasms, Glandular and Epithelial; Neoplastic Processes; Neoplasms, Neuroepithelial; Neuroectodermal Tumors; Neoplasms, Germ Cell and Embryonal; Neoplasms, Nerve Tissue; Neuroectodermal Tumors, Primitive, Peripheral; Neuroectodermal Tumors, Primitive; Neoplasms, Second Primary; Neoplasm Metastasis; Neuroblastoma
• Other Study ID Numbers: XH-24-008

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: YES
• IPD Plan Description: The investigators plan to share individual participant data (IPD) with other researchers to promote transparency and facilitate further research in the field. The shared data will include clinical outcomes, laboratory results, and demographic information.
• IPD Sharing Time Frame: The investigators anticipate that the data will be available for sharing six months after the primary study results have been published in a peer-reviewed journal and will remain accessible within one years after publication.

IPD Sharing Access Criteria

Access Plan for IPD:

Who will be able to access:

Researchers, academic institutions, and other scientists interested in conducting research in related fields.

What they will be able to access:

Clinical outcomes, laboratory test results, demographic information, and other supporting information related to the study.

How they will be able to access it:

They will need to submit an access request form, outlining the purpose of their research. All requests will be reviewed by the principal investigator of the study, who will determine whether access is granted based on research ethics and data use policies. Data will be provided through a secure online platform to ensure privacy and security.

• IPD Sharing Supporting Information Type: STUDY_PROTOCOL; SAP; ICF; ANALYTIC_CODE; CSR

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No
• product manufactured in and exported from the U.S.: No