Artificial intelligence-based morphologic classification and molecular characterization of neuroblastic tumors from digital histopathology.

A deep learning model using attention-based multiple instance learning (aMIL) and self-supervised learning (SSL) was developed to perform pathologic classification of neuroblastic tumors and assess MYCN-amplification status using H&E-stained whole slide digital images. The model demonstrated strong performance in identifying diagnostic category, grade, mitosis-karyorrhexis index (MKI), and MYCN-amplification on an external test dataset. This AI-based approach establishes a valuable tool for automating diagnosis and precise classification of neuroblastoma tumors.

Small-molecule inhibition of the METTL3/METTL14 complex suppresses neuroblastoma tumor growth and promotes differentiation.

The N6-methyladenosine (m6A) RNA modification is an important regulator of gene expression. m6A is deposited by a methyltransferase complex that includes methyltransferase-like 3 (METTL3) and methyltransferase-like 14 (METTL14). High levels of METTL3/METTL14 drive the growth of many types of adult cancer, and METTL3/METTL14 inhibitors are emerging as new anticancer agents. However, little is known about the m6A epitranscriptome or the role of the METTL3/METTL14 complex in neuroblastoma, a common pediatric cancer. Here, we show that METTL3 knockdown or pharmacologic inhibition with the small molecule STM2457 leads to reduced neuroblastoma cell proliferation and increased differentiation. These changes in neuroblastoma phenotype are associated with decreased m6A deposition on transcripts involved in nervous system development and neuronal differentiation, with increased stability of target mRNAs. In preclinical studies, STM2457 treatment suppresses the growth of neuroblastoma tumors in vivo. Together, these results support the potential of METTL3/METTL14 complex inhibition as a therapeutic strategy against neuroblastoma.

5-Hydroxymethylcytosine Profiling of Cell-Free DNA Identifies Bivalent Genes That Are Prognostic of Survival in High-Risk Neuroblastoma.

PURPOSE: Neuroblastoma is the most common extracranial solid tumor in childhood. We previously showed that circulating cell-free DNA (cfDNA) and tumor biopsy derived 5-hydroxymethylcytosime (5-hmC) profiles identified patients with neuroblastoma who experienced subsequent relapse. Here, we hypothesized that 5-hmC modifications selectively enriched in cfDNA compared with tumor biopsy samples would identify epigenetic changes associated with aggressive tumor behavior and identify novel biomarkers of outcome in patients with high-risk neuroblastoma. METHODS: 5-hmC profiles from cfDNA (n = 64) and tumor biopsies (n = 48) were compared. Two neuroblastoma cell lines underwent chromatin immunoprecipitation followed by sequencing (ChIP-Seq) for H3K27me3, H3K4me3, and H3K27ac; kethoxal-associated single-stranded DNA sequencing; hmC-Seal for 5-hmC; and RNA-sequencing (RNA-Seq). Genes enriched for both H3K27me3 and H3K4me3 in the included cell lines were defined as bivalent. Using bivalent genes defined in vitro, a bivalent signature was established in three publicly available cohorts of patients with neuroblastoma through gene set variation analysis. Differences between tumors with high or low bivalent signatures were assessed by the Kaplan-Meier method and Cox proportional hazards models. RESULTS: In cfDNA compared with tumor biopsy derived 5-hmC profiles, we found increased 5-hmC deposition on Polycomb Repressive Complex 2 target genes, a finding previously described in the context of bivalent genes. We identified 313 genes that bore bivalent chromatin marks, were enriched for mediators of neuronal differentiation, and were transcriptionally repressed across a panel of heterogeneous neuroblastoma cell lines. In three distinct clinical cohorts, low bivalent signature was significantly and independently associated with worse clinical outcome in patients with high-risk neuroblastoma. CONCLUSION: Low expression of bivalent genes is a biomarker of worse outcome in patients with high-risk neuroblastoma.

Diffuse Pediatric-type High-grade Glioma Arising in an Ovarian Mature Cystic Teratoma.

Immature neuroectodermal tissue can be found in the ovary as part of an immature teratoma or as part of a teratoma with malignant neuroectodermal transformation. Such lesions may closely resemble central nervous system tumors, but their biologic similarity is unclear. We describe an 18-yr-old female who presented with abdominal pain caused by an ovarian mass with widespread metastases. Histology showed a primitive, high-grade tumor arising in the background of a mature teratoma. The tumor was SOX10 positive, with focal expression of GFAP, S100, NSE, and synaptophysin. Molecular analysis demonstrated co-amplification of PDGFRA and KIT , alterations common in high-grade gliomas. By whole-genome methylation profiling, it clustered into the "diffuse pediatric-type high-grade glioma, RTK1 subtype, subclass c" group. Despite progressing through 2 lines of chemotherapy with widespread metastatic disease, she achieved an excellent response to chemotherapy directed toward aggressive germ cell tumors. This case emphasizes the importance of immunohistochemical, genomic, and epigenetic analyses to accurately classify these exceedingly rare tumors and determine the optimal therapy.

Adrenergic and mesenchymal signatures are identifiable in cell-free DNA and correlate with metastatic disease burden in children with neuroblastoma.

BACKGROUND: Cell-free DNA (cfDNA) profiles of 5-hydroxymethylcytosine (5-hmC), an epigenetic marker of open chromatin and active gene expression, are correlated with metastatic disease burden in patients with neuroblastoma. Neuroblastoma tumors are comprised of adrenergic (ADRN) and mesenchymal (MES) cells, and the relative abundance of each in tumor biopsies has prognostic implications. We hypothesized that ADRN and MES-specific signatures could be quantified in cfDNA 5-hmC profiles and would augment the detection of metastatic burden in patients with neuroblastoma. METHODS: We previously performed an integrative analysis to identify ADRN and MES-specific genes (n = 373 and n = 159, respectively). Purified DNA from cell lines was serial diluted with healthy donor cfDNA. Using Gene Set Variation Analysis (GSVA), ADRN and MES signatures were optimized. We then quantified signature scores, and our prior neuroblastoma signature, in cfDNA from 84 samples from 46 high-risk patients including 21 patients with serial samples. RESULTS: Samples from patients with higher metastatic burden had increased GSVA scores for both ADRN and MES gene signatures (p < .001). While ADRN and MES signature scores tracked together in serially collected samples, we identified instances of patients with increases in either MES or ADRN score at relapse. CONCLUSIONS: While it is feasible to identify ADRN and MES signatures using 5-hmC profiles of cfDNA from neuroblastoma patients and correlate these signatures to metastatic burden, additional data are needed to determine the optimal strategies for clinical implementation. Prospective evaluation in larger cohorts is ongoing.

Adrenergic and mesenchymal signatures are identifiable in cell-free DNA and correlate with metastatic disease burden in children with neuroblastoma.

Background: Cell free DNA (cfDNA) profiles of 5-hydroxymethylcytosine (5-hmC), an epigenetic marker of open chromatin and active gene expression, are correlated with metastatic disease burden in patients with neuroblastoma. Neuroblastoma tumors are comprised of adrenergic (ADRN) and mesenchymal (MES) cells, and the relative abundance of each in tumor biopsies has prognostic implications. We hypothesized that ADRN and MES specific signatures could be quantified in cfDNA 5-hmC profiles and would augment the detection of metastatic burden in patients with neuroblastoma. Methods: We previously performed an integrative analysis to identify ADRN and MES specific genes (n=373 and n=159, respectively). Purified DNA from cell lines was serial diluted with healthy donor cfDNA. Using Gene Set Variation Analysis (GSVA), ADRN and MES signatures were optimized. We then quantified signature scores, and our prior neuroblastoma signature, in cfDNA from 84 samples from 46 high-risk patients including 21 patients with serial samples. Results: Samples from patients with higher metastatic burden had increased GSVA scores for both ADRN and MES gene signatures (p < 0.001). While ADRN and MES signature scores tracked together in serially collected samples, we identified instances of patients with increases in either MES or ADRN score at relapse. Conclusions: While it is feasible to identify ADRN and MES signatures using 5-hmC profiles of cfDNA from neuroblastoma patients and correlate these signatures to metastatic burden, additional data are needed to determine the optimal strategies for clinical implementation. Prospective evaluation in larger cohorts is ongoing.

T-cell inflammation is prognostic of survival in patients with high-risk neuroblastoma enriched for an adrenergic signature.

Purpose: T-cell inflammation (TCI) has been shown to be a prognostic marker in neuroblastoma, a tumor comprised of cells that can exist in two epigenetic states, adrenergic (ADRN) and mesenchymal (MES). We hypothesized that elucidating unique and overlapping aspects of these biologic features could serve as novel biomarkers. Patients and Methods: We detected lineage-specific, single-stranded super-enhancers defining ADRN and MES specific genes. Publicly available neuroblastoma RNA-seq data from GSE49711 (Cohort 1) and TARGET (Cohort 2) were assigned MES, ADRN, and TCI scores. Tumors were characterized as MES (top 33%) or ADRN (bottom 33%), and TCI (top 67% TCI score) or non-inflamed (bottom 33% TCI score). Overall survival (OS) was assessed using the Kaplan-Meier method, and differences were assessed by the log-rank test. Results: We identified 159 MES genes and 373 ADRN genes. TCI scores were correlated with MES scores (R=0.56, p<0.001 and R=0.38, p<0.001) and anticorrelated with MYCN -amplification (R=-0.29, p<0.001 and -0.18, p=0.03) in both cohorts. Among Cohort 1 patients with high-risk, ADRN tumors (n=59), those with TCI tumors (n=22) had superior OS to those with non-inflammed tumors (n=37) (p=0.01), though this comparison did not reach significance in Cohort 2. TCI status was not associated with survival in patients with high-risk MES tumors in either cohort. Conclusions: High inflammation scores were correlated with improved survival in some high-risk patients with, ADRN but not MES neuroblastoma. These findings have implications for approaches to treating high-risk neuroblastoma.

Histotripsy induces apoptosis and reduces hypoxia in a neuroblastoma xenograft model.

BACKGROUND: Neuroblastoma (NB) is the most common extracranial solid tumor of childhood, and high-risk disease is resistant to intensive treatment. Histotripsy is a focused ultrasound therapy under development for tissue ablation via bubble activity. The goal of this study was to assess outcomes of histotripsy ablation in a xenograft model of high-risk NB. METHODS: Female NCr nude mice received NGP-luciferase cells intrarenally. Under ultrasound image guidance, histotripsy pulses were applied over a distance of 4-6 mm within the tumors. Bioluminescence indicative of tumor viability was quantified before, immediately after, and 24 h after histotripsy exposure. Tumors were immunostained to assess apoptosis (TUNEL), endothelium (endomucin), pericytes (αSMA), hypoxia (pimonidazole), vascular endothelial growth factor A (VEGFA), and platelet-derived growth factor-B (PDGF-B). The apoptotic cytokine TNFα and its downstream effector cleaved caspase-3 (c-casp-3) were assessed with SDS-PAGE. RESULTS: Histotripsy induced a 50% reduction in bioluminescence compared to untreated controls, with an absence of nuclei in the treatment core surrounded by a dense rim of TUNEL-positive cells. Tumor regions not targeted by histotripsy also showed an increase in TUNEL staining density. Increased apoptosis in histotripsy samples was consistent with increases in TNFα and c-casp-3 relative to controls. Treated tumors exhibited a decrease in hypoxia, VEGF, PDGF-B, and pericyte coverage of vasculature compared to control samples. Further, increases in vasodilation were found in histotripsy-treated specimens. CONCLUSIONS: In addition to ablative effects, histotripsy was found to drive tumor apoptosis through intrinsic pathways, altering blood vessel architecture, and reducing hypoxia.

5-hydroxymethylcytosine profiling of cell-free DNA identifies bivalent genes that are prognostic of survival in high-risk neuroblastoma.

Neuroblastoma is the most common extra-cranial solid tumor in childhood and epigenetic dysregulation is a key driver of this embryonal disease. In cell-free DNA from neuroblastoma patients with high-risk disease, we found increased 5-hydroxymethylcytosine (5-hmC) deposition on Polycomb Repressive Complex 2 (PRC2) target genes, a finding previously described in the context of bivalent genes. As bivalent genes, defined as genes bearing both activating (H3K4me3) and repressive (H3K27me3) chromatin modifications, have been shown to play an important role in development and cancer, we investigated the potential role of bivalent genes in maintaining a de-differentiated state in neuroblastoma and their potential use as a biomarker. We identified 313 genes that bore bivalent chromatin marks, were enriched for mediators of neuronal differentiation, and were transcriptionally repressed across a panel of heterogenous neuroblastoma cell lines. Through gene set variance analysis, we developed a clinically implementable bivalent signature. In three distinct clinical cohorts, low bivalent signature was significantly and independently associated with worse clinical outcome in high-risk neuroblastoma patients. Thus, low expression of bivalent genes is a biomarker of ultra-high-risk disease and may represent a therapeutic opportunity in neuroblastoma.

Methyltransferase Inhibition Enables Tgfß Driven Induction of CDKN2A and B in Cancer Cells.

CDKN2A/B deletion or silencing is common across human cancer, reinforcing the general importance of bypassing its tumor suppression in cancer formation or progression. In rhabdomyosarcoma (RMS) and neuroblastoma, two common childhood cancers, the three CDKN2A/B transcripts are independently expressed to varying degrees, but one, ARF, is uniformly silenced. Although TGFß induces certain CDKN2A/B transcripts in HeLa cells, it was unable to do so in five tested RMS lines unless the cells were pretreated with a broadly acting methyltransferase inhibitor, DZNep, or one targeting EZH2. CDKN2A/B induction by TGFß correlated with de novo appearance of three H3K27Ac peaks within a 20 kb cis element ∼150 kb proximal to CDKN2A/B. Deleting that segment prevented their induction by TGFß but not a basal increase driven by methyltransferase inhibition alone. Expression of two CDKN2A/B transcripts was enhanced by dCas9/CRISPR activation targeting either the relevant promoter or the 20 kb cis elements, and this "precise" manipulation diminished RMS cell propagation in vitro. Our findings show crosstalk between methyltransferase inhibition and TGFß-dependent activation of a remote enhancer to reverse CDKN2A/B silencing. Though focused on CDKN2A/B here, such crosstalk may apply to other TGFß-responsive genes and perhaps govern this signaling protein's complex effects promoting or blocking cancer.

Clinical Targeted Next-Generation Panel Sequencing Reveals MYC Amplification Is a Poor Prognostic Factor in Osteosarcoma.

PURPOSE: Osteosarcoma risk stratification, on the basis of the presence of metastatic disease at diagnosis and histologic response to chemotherapy, has remained unchanged for four decades, does not include genomic features, and has not facilitated treatment advances. We report on the genomic features of advanced osteosarcoma and provide evidence that genomic alterations can be used for risk stratification. MATERIALS AND METHODS: In a primary analytic patient cohort, 113 tumor and 69 normal samples from 92 patients with high-grade osteosarcoma were sequenced with OncoPanel, a targeted next-generation sequencing assay. In this primary cohort, we assessed the genomic landscape of advanced disease and evaluated the correlation between recurrent genomic events and outcome. We assessed whether prognostic associations identified in the primary cohort were maintained in a validation cohort of 86 patients with localized osteosarcoma tested with MSK-IMPACT. RESULTS: In the primary cohort, 3-year overall survival (OS) was 65%. Metastatic disease, present in 33% of patients at diagnosis, was associated with poor OS (P = .04). The most frequently altered genes in the primary cohort were TP53, RB1, MYC, CCNE1, CCND3, CDKN2A/B, and ATRX. Mutational signature 3 was present in 28% of samples. MYC amplification was associated with a worse 3-year OS in both the primary cohort (P = .015) and the validation cohort (P = .012). CONCLUSION: The most frequently occurring genomic events in advanced osteosarcoma were similar to those described in prior reports. MYC amplification, detected with clinical targeted next-generation sequencing panel tests, is associated with poorer outcomes in two independent cohorts.

Breaking Up Isn't Hard to Do: Isolating Cell-free DNA Fragments in Osteosarcoma.

The novel use of blood-based biospecimens from a retrospective cohort of 50 patients with osteosarcoma was recently studied. The potential clinical utility of sorting cell-free DNA by fragment size was defined, with shorter tumor-specific DNA enrichment providing prognostic value and allowing for streamlined molecular profiling of circulating tumor material. See related article by Udomruk et al., p. 2085.

Persistence of Racial and Ethnic Disparities in Risk and Survival for Patients with Neuroblastoma over Two Decades.

BACKGROUND: Racial/ethnic survival disparities in neuroblastoma were first reported more than a decade ago. We sought to investigate if these disparities have persisted with current era therapy. METHODS: Two patient cohorts were identified in the International Neuroblastoma Risk Group Data Commons (INRGdc) (Cohort 1: diagnosed 2001-2009, n=4359; Cohort 2: diagnosed 2010-2019, n=4891). Chi-squared tests were used to assess the relationship between race/ethnicity and clinical and biologic features. Survival was estimated by the Kaplan-Meier method. Cox proportional hazards regression analyses were performed to investigate the association between racial/ethnic groups and prognostic markers. RESULTS: Significantly higher 5-year event-free survival (EFS) and overall survival (OS) were observed for Cohort 2 compared to Cohort 1 (P<0.001 and P<0.001, respectively). Compared to White patients, Black patients in both cohorts had a higher proportion of high-risk disease (Cohort 1: P<0.001; Cohort 2: P<0.001) and worse EFS (Cohort 1: P<0.001; Cohort 2 P<0.001) and OS (Cohort 1: P<0.001; Cohort 2: P<0.001). In Cohort 1, Native Americans also had a higher proportion of high-risk disease (P=0.03) and inferior EFS/OS. No significant survival disparities were observed for low- or intermediate-risk patients in either cohort or high-risk patients in Cohort 1. Hispanic patients with high-risk disease in Cohort 2 had significantly inferior OS (P=0.047). Significantly worse OS, but not EFS, (P=0.006 and P=0.02, respectively) was also observed among Black and Hispanic patients assigned to receive post-Consolidation dinutuximab on clinical trials (n=885). CONCLUSION: Racial/ethnic survival disparities have persisted over time and were observed among high-risk patients assigned to receive post-Consolidation dinutuximab.

Systemic Bevacizumab for Severe Recurrent Respiratory Papillomatosis.

Recurrent respiratory papillomatosis (RRP) is the most common benign pediatric laryngeal neoplasm. Various adjuvant medical therapies have failed to reliably decrease surgical frequency in this challenging airway disease. Recently, systemic bevacizumab has shown promise in advanced, treatment-resistant papillomatosis. We describe the use of systemic bevacizumab in two children with severe RRP unresponsive to other therapies. Voice and breathing improved dramatically in both patients with minimal side effects. Both patients have not required surgery in 24 months and 16 months, respectively. Systemic bevacizumab is a promising long-term treatment for severe RRP, with oncology playing an important role in patient care.

MYCN and HIF-1 directly regulate TET1 expression to control 5-hmC gains and enhance neuroblastoma cell migration in hypoxia.

Ten-Eleven-Translocation 5-methylcytosine dioxygenases 1-3 (TET1-3) convert 5-methylcytosine to 5-hydroxymethylcytosine (5-hmC), using oxygen as a co-substrate. Contrary to expectations, hypoxia induces 5-hmC gains in MYCN-amplified neuroblastoma (NB) cells via upregulation of TET1. Here, we show that MYCN directly controls TET1 expression in normoxia, and in hypoxia, HIF-1 augments TET1 expression and TET1 protein stability. Through gene-editing, we identify two MYCN and HIF-1 binding sites within TET1 that regulate gene expression. Bioinformatic analyses of 5-hmC distribution and RNA-sequencing data from hypoxic cells implicate hypoxia-regulated genes important for cell migration, including CXCR4. We show that hypoxic cells lacking the two MYCN/HIF-1 binding sites within TET1 migrate slower than controls. Treatment of MYCN-amplified NB cells with a CXCR4 antagonist results in slower migration under hypoxic conditions, suggesting that inclusion of a CXCR4 antagonist into NB treatment regimens could be beneficial for children with MYCN-amplified NBs.

In MYCN-amplified neuroblastoma cell lines, MYCN directly controls TET1 expression in normoxia.In MYCN-amplified neuroblastoma cell lines exposed to hypoxia, HIF-1 augments TET1 expression and TET1 protein stability.Hypoxic MYCN-amplified neuroblastoma cell lines have increased cell migration, mediated by genes including CXCR4 that gain 5-hydroxymethylcytosine density.Treatment of MYCN-amplified NB cells with a CXCR4 antagonist slows hypoxia-associated migration, suggesting a CXCR4 antagonist could be beneficial in treatment regimens for children with MYCN-amplified neuroblastomas.

Efficacy of post-induction therapy for high-risk neuroblastoma patients with end-induction residual disease.

BACKGROUND: High-risk neuroblastoma patients with end-induction residual disease commonly receive post-induction therapy in an effort to increase survival by improving the response before autologous stem cell transplantation (ASCT). The authors conducted a multicenter, retrospective study to investigate the efficacy of this approach. METHODS: Patients diagnosed between 2008 and 2018 without progressive disease with a partial response or worse at end-induction were stratified according to the post-induction treatment: 1) no additional therapy before ASCT (cohort 1), 2) post-induction "bridge" therapy before ASCT (cohort 2), and 3) post-induction therapy without ASCT (cohort 3). χ2 tests were used to compare patient characteristics. Three-year event-free survival (EFS) and overall survival (OS) were estimated by the Kaplan-Meier method and survival curves were compared by log-rank test. RESULTS: The study cohort consisted of 201 patients: cohort 1 (n = 123), cohort 2 (n = 51), and cohort 3 (n = 27). Although the end-induction response was better for cohort 1 than cohorts 2 and 3, the outcomes for cohorts 1 and 2 were not significantly different (P = .77 for EFS and P = .85 for OS). Inferior outcomes were observed for cohort 3 (P < .001 for EFS and P = .06 for OS). Among patients with end-induction stable metastatic disease, 3-year EFS was significantly improved for cohort 2 versus cohort 1 (P = .04). Cohort 3 patients with a complete response at metastatic sites after post-induction therapy had significantly better 3-year EFS than those with residual metastatic disease (P = .01). CONCLUSIONS: Prospective studies to confirm the benefits of bridge treatment and the prognostic significance of metastatic response observed in this study are warranted.

Molecular profiling identifies targeted therapy opportunities in pediatric solid cancer.

To evaluate the clinical impact of molecular tumor profiling (MTP) with targeted sequencing panel tests, pediatric patients with extracranial solid tumors were enrolled in a prospective observational cohort study at 12 institutions. In the 345-patient analytical population, median age at diagnosis was 12 years (range 0-27.5); 298 patients (86%) had 1 or more alterations with potential for impact on care. Genomic alterations with diagnostic, prognostic or therapeutic significance were present in 61, 16 and 65% of patients, respectively. After return of the results, impact on care included 17 patients with a clarified diagnostic classification and 240 patients with an MTP result that could be used to select molecularly targeted therapy matched to identified alterations (MTT). Of the 29 patients who received MTT, 24% had an objective response or experienced durable clinical benefit; all but 1 of these patients received targeted therapy matched to a gene fusion. Of the diagnostic variants identified in 209 patients, 77% were gene fusions. MTP with targeted panel tests that includes fusion detection has a substantial clinical impact for young patients with solid tumors.

Predicting Response to Chemotherapy in Patients With Newly Diagnosed High-Risk Neuroblastoma: A Report From the International Neuroblastoma Risk Group.

PURPOSE: Metaiodobenzylguanidine (MIBG) scans are a radionucleotide imaging modality that undergo Curie scoring to semiquantitatively assess neuroblastoma burden, which can be used as a marker of therapy response. We hypothesized that a convolutional neural network (CNN) could be developed that uses diagnostic MIBG scans to predict response to induction chemotherapy. METHODS: We analyzed MIBG scans housed in the International Neuroblastoma Risk Group Data Commons from patients enrolled in the Children's Oncology Group high-risk neuroblastoma study ANBL12P1. The primary outcome was response to upfront chemotherapy, defined as a Curie score ≤ 2 after four cycles of induction chemotherapy. We derived and validated a CNN using two-dimensional whole-body MIBG scans from diagnosis and evaluated model performance using area under the receiver operating characteristic curve (AUC). We also developed a clinical classification model to predict response on the basis of age, stage, and MYCN amplification. RESULTS: Among 103 patients with high-risk neuroblastoma included in the final cohort, 67 (65%) were responders. Performance in predicting response to upfront chemotherapy was equivalent using the CNN and the clinical model. Class-activation heatmaps verified that the CNN used areas of disease within the MIBG scans to make predictions. Furthermore, integrating predictions using a geometric mean approach improved detection of responders to upfront chemotherapy (geometric mean AUC 0.73 v CNN AUC 0.63, P < .05; v clinical model AUC 0.65, P < .05). CONCLUSION: We demonstrate feasibility in using machine learning of diagnostic MIBG scans to predict response to induction chemotherapy for patients with high-risk neuroblastoma. We highlight improvements when clinical risk factors are also integrated, laying the foundation for using a multimodal approach to guiding treatment decisions for patients with high-risk neuroblastoma.

Germline Sequencing Improves Tumor-Only Sequencing Interpretation in a Precision Genomic Study of Patients With Pediatric Solid Tumor.

PURPOSE: Molecular tumor profiling is becoming a routine part of clinical cancer care, typically involving tumor-only panel testing without matched germline. We hypothesized that integrated germline sequencing could improve clinical interpretation and enhance the identification of germline variants with significant hereditary risks. MATERIALS AND METHODS: Tumors from pediatric patients with high-risk, extracranial solid malignancies were sequenced with a targeted panel of cancer-associated genes. Later, germline DNA was analyzed for a subset of these genes. We performed a post hoc analysis to identify how an integrated analysis of tumor and germline data would improve clinical interpretation. RESULTS: One hundred sixty participants with both tumor-only and germline sequencing reports were eligible for this analysis. Germline sequencing identified 38 pathogenic or likely pathogenic variants among 35 (22%) patients. Twenty-five (66%) of these were included in the tumor sequencing report. The remaining germline pathogenic or likely pathogenic variants were single-nucleotide variants filtered out of tumor-only analysis because of population frequency or copy-number variation masked by additional copy-number changes in the tumor. In tumor-only sequencing, 308 of 434 (71%) single-nucleotide variants reported were present in the germline, including 31% with suggested clinical utility. Finally, we provide further evidence that the variant allele fraction from tumor-only sequencing is insufficient to differentiate somatic from germline events. CONCLUSION: A paired approach to analyzing tumor and germline sequencing data would be expected to improve the efficiency and accuracy of distinguishing somatic mutations and germline variants, thereby facilitating the process of variant curation and therapeutic interpretation for somatic reports, as well as the identification of variants associated with germline cancer predisposition.

Applications of Artificial Intelligence in Pediatric Oncology: A Systematic Review.

PURPOSE: There is a need for an improved understanding of clinical and biologic risk factors in pediatric cancer to improve patient outcomes. Machine learning (ML) represents the application of computational inference from advanced statistical methods that can be applied to increasing amount of data available for study in pediatric oncology. The goal of this systematic review was to systematically characterize the state of ML in pediatric oncology and highlight advances and opportunities in the field. METHODS: We conducted a systematic review of the Embase, Scopus, and MEDLINE databases for applications of ML in pediatric oncology. Query results from all three databases were aggregated and duplicate studies were removed. RESULTS: A total of 42 unique articles that examined the applications of ML in pediatric oncology met inclusion criteria for review. We identified 20 studies of CNS tumors, 13 of solid tumors, and nine of leukemia. ML tasks included classification, prediction of treatment response, and dose optimization with a variety of methods being used including neural network, k-nearest neighbor, random forest, naive Bayes, and support vector machines. Strengths of the identified studies included matching or outperforming physician comparators via automated analysis and predicting therapeutic response. Common limitations included significant heterogeneity in reporting standards, clinical applicability, small sample sizes, and missing external validation cohorts. CONCLUSION: We identified areas where ML can enhance clinical care in ways that may not otherwise be achievable. Although ML promises enormous potential in improving diagnostics, decision making, and monitoring for children with cancer, the field remains in early stages and future work will be aided by standards and guidelines to ensure rigorous methodologic design and maximizing clinical utility.