Immunogenomic determinants of tumor microenvironment correlate with survival in high-risk neuroblastoma

Background:

Tumor-infiltrating CD8+ T cells and neoantigens are predictors of a favorable prognosis and response to immunotherapy in adult cancer, but little is known about their role in pediatric malignancies. Here, we analyzed the prognostic strength of T cell-inflamed (Tinfl) expression and neoantigen load (neo) in high-risk neuroblastoma (NBL). We investigated transcriptional programs associated with the non-T cell-inflamed (non-Tinfl) phenotype for mechanisms of immune exclusion.

Methods:

Using a defined Tinfl gene expression signature, we categorized high-risk NBL from the TARGET (Therapeutically Applicable Research to Generate Effective Treatments, n=123) and the GMKF (Gabriella Miller Kids First, n=48) programs into non-Tinfl, Tinfl, and intermediate groups. Cox proportional hazard models were used to test the association between survival and tumor group, neo, MYCN amplification. Pathways activated in non-Tinfl vs. Tinfl tumors were identified.

Results:

Patients with high-risk Tinfl tumors showed improved overall survival (OS) compared to those with non-Tinfl tumors (P<0.05), independent of MYCN amplification, in both TARGET and GMKF. Higher neo was associated with better event-free survival and OS (P<0.005) and was independent of Tinfl. Activation of MYCN, ASCL1, SOX11, and KMT2A was inversely correlated with the Tinfl signature in both cohorts.

Conclusions:

Our results indicate that tumors from children with high-risk NBL harboring a strong Tinfl signature have a favorable clinical outcome, and neo is a prognosis predictor, independent of Tinfl. Strategies to target SOX11 and other signaling pathways associated with non-Tinfl tumors should be pursued as potential immune-potentiating interventions.

A nomogram of clinical and biologic factors to predict survival in children newly diagnosed with high-risk neuroblastoma: An International Neuroblastoma Risk Group project

BACKGROUND

Long-term outcome remains poor for children with high-risk neuroblastoma (five-year overall survival [OS] ∼50%). Our objectives were to (a) identify prognostic biomarkers and apply them in a nomogram to identify the subgroup of ultra-high-risk patients at highest risk of disease progression/death, for whom novel frontline therapy is urgently needed; and (b) validate the nomogram in an independent cohort.

METHODS

A total of 1820 high-risk patients (≥18 months old with metastatic neuroblastoma), diagnosed 1998-2015, from the International Neuroblastoma Risk Groups (INRG) Data Commons were analyzed in a retrospective cohort study. Using multivariable Cox regression of OS from diagnosis, a nomogram was created from prognostic biomarkers to predict three-year OS. External validation was performed using the SIOPEN HR-NBL1 trial cohort (n = 521), evidenced by receiver operating characteristic curves.

RESULTS

The nomogram, including MYCN status (P < 0.0001), lactate dehydrogenase (LDH) (P = 0.0007), and presence of bone marrow metastases (P = 0.004), had robust performance and was validated. Applying the nomogram at diagnosis (a) gives prognosis of an individual patient and (b) identifies patients predicted to have poor outcome (three-year OS was 30% ± 5% for patients with a nomogram score of > 82 points; 58% ± 1% for those ≤82 points). Median follow-up time was 5.5 years (range, 0-14.1).

CONCLUSIONS

In high-risk neuroblastoma, a novel, publicly available nomogram using prognostic biomarkers (MYCN status, LDH, presence of bone marrow metastases; https://neuroblastoma.shinyapps.io/High-Risk-Neuroblastoma-Nomogram/) has the flexibility to apply a clinically suitable and context-specific cutoff to identify patients at highest risk of death. This will facilitate testing urgently needed new frontline treatment options to improve outcome for these children.

Long-Term Follow-up of a Phase III Study of ch14.18 (Dinutuximab) + Cytokine Immunotherapy in Children with High-Risk Neuroblastoma: COG Study ANBL0032

PURPOSE

Previously our randomized phase III trial demonstrated that immunotherapy including dinutuximab, a chimeric anti-GD2 mAb, GM-CSF, and IL2 improved survival for children with high-risk neuroblastoma that had responded to induction and consolidation therapy. These results served as the basis for FDA approval of dinutuximab. We now present long-term follow-up results and evaluation of predictive biomarkers.

PATIENTS AND METHODS

Patients recieved six cycles of isotretinoin with or without five cycles of immunotherapy which consists of dinutuximab with GM-CSF alternating with IL2. Accrual was discontinued early due to meeting the protocol-defined stopping rule for efficacy, as assessed by 2-year event-free survival (EFS). Plasma levels of dinutuximab, soluble IL2 receptor (sIL2R), and human anti-chimeric antibody (HACA) were assessed by ELISA. Fcγ receptor 2A and 3A genotypes were determined by PCR and direct sequencing.

RESULTS

For 226 eligible randomized patients, 5-year EFS was 56.6 ± 4.7% for patients randomized to immunotherapy (n = 114) versus 46.1 ± 5.1% for those randomized to isotretinoin only (n = 112; P = 0.042). Five-year overall survival (OS) was 73.2 ± 4.2% versus 56.6 ± 5.1% for immunotherapy and isotretinoin only patients, respectively (P = 0.045). Thirteen of 122 patients receiving dinutuximab developed HACA. Plasma levels of dinutuximab, HACA, and sIL2R did not correlate with EFS/OS, or clinically significant toxicity. Fcγ receptor 2A and 3A genotypes did not correlate with EFS/OS.

CONCLUSIONS

Immunotherapy with dinutuximab improved outcome for patients with high-risk neuroblastoma. Early stoppage for efficacy resulted in a smaller sample size than originally planned, yet clinically significant long-term differences in survival were observed.

Association between end-induction response according to the revised International Neuroblastoma Response Criteria (INRC) and outcome in high-risk neuroblastoma patients

BACKGROUND

The 1993 International Neuroblastoma Response Criteria (INRC) were revised in 2017 to include modern functional imaging studies and methods for quantifying disease in bone marrow. We hypothesized the 2017 INRC would enable more precise assessment of response to treatment and provide superior prognostic information compared with the 1993 criteria.

METHODS

High-risk (HR) neuroblastoma patients from two institutions in Chicago diagnosed between 2006 and 2016 were identified. Patients were assessed post induction chemotherapy via the 1993 and 2017 INRC and classified as responder (≥ mixed response [MXR] or ≥ minor response [MR], respectively) or nonresponder (< MXR or < MR). Event-free survival (EFS) and overall survival (OS) for responders versus nonresponders were determined from end induction and stratified by Cox regression. Patients with progressive disease at end induction were eliminated from the EFS analyses but included in the OS analysis.

RESULTS

The 1993 criteria classified 52 of the 60 HR patients as responders, whereas 54 responders were identified using the 2017 criteria (Spearman correlation r = 0.82, P < 0.001). No statistically significant difference in EFS was observed for responders versus nonresponders using either criteria (P = 0.48 and P = 0.08). However, superior OS was observed for responders (P = 0.01) using either criteria. Both criteria were sensitive in identifying responders among those with good outcomes. The specificity to identify nonresponders among those with poor outcomes was poor.

CONCLUSIONS

In HR neuroblastoma, end-induction response defined by the 1993 or 2017 INRC is associated with survival. Larger cohorts are needed to determine if the 2017 INRC provides more precise prognostication.

Racial and Ethnic Differences in Communication and Care for Children With Advanced Cancer

CONTEXT

Racial and ethnic disparities in end-of-life care are well documented among adults with advanced cancer.

OBJECTIVES

To examine the extent to which communication and care differ by race and ethnicity among children with advanced cancer.

METHODS

We conducted a prospective cohort study at nine pediatric cancer centers enrolling 95 parents (42% racial/ethnic minorities) of children with poor prognosis cancer (relapsed/refractory high-risk neuroblastoma). Parents were surveyed about whether prognosis was discussed; likelihood of cure; intent of current treatment; and primary goal of care. Medical records were used to identify high-intensity medical care since the most recent recurrence. Logistic regression evaluated differences between white non-Hispanic and minority (black, Hispanic, and Asian/other race) parents.

RESULTS

About 26% of parents recognized the child's low likelihood of cure. Minority parents were less likely to recognize the poor prognosis (odds ratio [OR] = 0.19; 95% CI = 0.06-0.63; P = 0.006) and the fact that current treatment was unlikely to offer cure (OR = 0.07; 95% CI = 0.02-0.27; P < 0.0001). Children of minority parents were more likely to experience high-intensity medical care (OR = 3.01; 95% CI = 1.29-7.02; P = 0.01). After adjustment for understanding of prognosis, race/ethnicity was no longer associated with high-intensity medical care (adjusted odds ratio = 2.14; 95% CI = 0.84-5.46; P = 0.11), although power to detect an association was limited.

CONCLUSION

Parental understanding of prognosis is limited across racial and ethnic groups; racial and ethnic minorities are disproportionately affected. Perhaps as a result, minority children experience higher rates of high-intensity medical care. Work to improve prognostic understanding should include focused work to meet needs of minority populations.

Tailoring Therapy for Children With Neuroblastoma on the Basis of Risk Group Classification: Past, Present, and Future

For children with neuroblastoma, the likelihood of cure varies widely according to age at diagnosis, disease stage, and tumor biology. Treatments are tailored for children with this clinically heterogeneous malignancy on the basis of a combination of markers that are predictive of risk of relapse and death. Sequential risk-based, cooperative-group clinical trials conducted during the past 4 decades have led to improved outcome for children with neuroblastoma. Increasingly accurate risk classification and refinements in treatment stratification strategies have been achieved with the more recent discovery of robust genomic and molecular biomarkers. In this review, we discuss the history of neuroblastoma risk classification in North America and Europe and highlight efforts by the International Neuroblastoma Risk Group (INRG) Task Force to develop a consensus approach for pretreatment stratification using seven risk criteria including an image-based staging system-the INRG Staging System. We also update readers on the current Children's Oncology Group risk classifier and outline plans for the development of a revised 2021 Children's Oncology Group classifier that will incorporate INRG Staging System criteria to facilitate harmonization of risk-based frontline treatment strategies conducted around the globe. In addition, we discuss new approaches to establish increasingly robust, future risk classification algorithms that will further refine treatment stratification using machine learning tools and expanded data from electronic health records and the INRG Data Commons.

The prognostic strength of serum LDH and serum ferritin in children with neuroblastoma: A report from the International Neuroblastoma Risk Group (INRG) project

PURPOSE

Age, MYCN status, stage, and histology have been used as neuroblastoma (NB) risk factors for decades. Serum lactate dehydrogenase (LDH) and serum ferritin are reproducible, easily obtained, and prognostic, though never used in risk stratification, except one German trial. We analyzed the prognostic strength of LDH and ferritin, overall, within high-risk NB, and by era, using the International Neuroblastoma Risk Group Data Commons.

PATIENTS AND METHODS

Children with NB (1990-2016) were categorized into LDH (n = 8867) and ferritin (n = 8575) risk groups using EFS. Cox models compared the prognostic strength of LDH and ferritin to age, MYCN status, and INSS stage.

RESULTS

Higher LDH conferred worse EFS, overall (5-year EFS) (100-899 IU/L: 76 ± 0.6%; 0-99 or 900-1399 IU/L: 60 ± 1.2%; ≥1400 IU/L: 36 ± 1.2%; P < .0001), and in high-risk NB post-2009 (3-year EFS) (117-381 IU/L: 67 ± 8.9%; 382-1334 IU/L: 58 ± 4.4%; 0-116 or ≥1335 IU/L: 46 ± 3.9%; P = .003). Higher ferritin conferred worse EFS, overall (5-year EFS) (1-29 ng/mL: 87 ± 0.9%; 0 or 30-89 ng/mL: 74 ± 0.8%; ≥90 ng/mL: 48 ± 0.9%; P < .0001), and in high-risk NB post-2009 (3-year EFS) (1-53 ng/mL: 71 ± 9.3%; 0 or 54-354 ng/mL: 55 ± 4.7%; ≥355 ng/mL: 34 ± 6.1%; P = .0008). In multivariable analyses adjusting for age, MYCN, and stage, LDH and ferritin maintained independent prognostic ability (P < .0001; adjusted HRs (95% CI): 1.7 (1.5-1.9), 2.3 (2.0-2.7), respectively).

CONCLUSIONS

LDH and ferritin are strongly prognostic in NB, overall and within high-risk NB patients treated post-2009 with modern therapy. LDH and ferritin show promise for (a) identifying ultra-high-risk; (b) refining risk stratification; and (c) clinical utility in low-/middle-income countries. Routine collection of LDH and ferritin should be reinitiated for evolving NB risk stratification.

Paraneoplastic opsoclonus myoclonus syndrome associated with inflammatory myofibroblastic tumor in a pediatric patient

Opsoclonus myoclonus syndrome (OMS) is a rare neurological syndrome caused by a paraneoplastic autoimmune process that affects children with neuroblastic tumors. Treatment includes corticosteroids, intravenous gamma globulin (IVIG), rituximab, and other immunosuppressive therapies. Here, we describe a patient diagnosed with OMS associated with a localized inflammatory myofibroblastic tumor. The patient has no evidence of tumor recurrence following surgical resection with 8-month follow-up. The neurologic symptoms resolved with corticosteroids and IVIG. This case demonstrates that in children, neoplasms other than neuroblastoma may be associated with this paraneoplastic syndrome, and highlights the importance of evaluating patients with OMS for underlying malignancies.

Abstract B02: 5-Hydroxymethylcytosine profiles in circulating cell-free DNA are biomarkers of disease burden in children with neuroblastoma

Background: 5-Hydroxymethylcytosine (5-hmC) is an epigenetic marker of open chromatin and active gene expression. The level of 5-hmC expression serves as a biomarker in adults with cancer, and we recently demonstrated the prognostic significance of tumor whole-genome 5-hmC profiling in children with neuroblastoma. Here, we used Nano-hmC-Seal technology to profile 5-hmC in circulating cell-free DNA (cfDNA) in blood samples from patients with neuroblastoma and age-matched healthy children.

Methods: Plasma was collected from 32 children with neuroblastoma at various time points during therapy and 34 well children. 5-hmC sequencing was performed using 100ng of cfDNA extracted from plasma. Metastatic disease burden was classified as high, moderate, low, or none based on Curie MIBG scores, the percentage of tumor cells in bone marrow samples, and/or measurable disseminated soft tissue disease.

Results: One hundred cfDNA 5-hmC profiles were generated from 27 high-risk, 3 intermediate-risk, and 2 low-risk patients and 34 well controls. Hierarchical clustering based on 5-hmC levels of 347 genes separated the samples into three main clusters corresponding to high, moderate, and low/no disease burden in the bone, bone marrow, and/or soft tissue. Using functional gene pathways, two subsets of patients within the low/no cluster were revealed that correlated with low tumor burden or no clinically detectable metastatic disease (sensitivity 70%, specificity 76.9%). 5-hmC profiles tracked closely with changes in disease status in 15 patients with serial samples. Samples from patients with a detectable tumor burden had genes with elevated 5-hmC in pathways of neuronal stem cell maintenance and the epigenetic regulatory complexes PRC2 and CTCF/cohesion.

Conclusions: cfDNA 5-hmC profiles correlate with disease burden in children with neuroblastoma and may serve as sensitive biomarkers of treatment response. Analysis of transcriptional networks regulated by these epigenomic modifications may lead to a deeper understanding of the pathways that drive resistance to treatment.

Citation Format: Mark A. Applebaum, Erin K. Barr, Jason Karpus, Meritxell Oliva, Alexandre Chlenski, Helen R. Salwen, Emma Wilkinson, Marija Dobratic, Elizabeth A. Sokol, Barbara E. Stranger, Chuan He, Susan L. Cohn. 5-Hydroxymethylcytosine profiles in circulating cell-free DNA are biomarkers of disease burden in children with neuroblastoma [abstract]. In: Proceedings of the AACR Special Conference on the Advances in Pediatric Cancer Research; 2019 Sep 17-20; Montreal, QC, Canada. Philadelphia (PA): AACR; Cancer Res 2020;80(14 Suppl):Abstract nr B02.

Age, Diagnostic Category, Tumor Grade, and Mitosis-Karyorrhexis Index Are Independently Prognostic in Neuroblastoma: An INRG Project

PURPOSE

The Children's Oncology Group (COG) stratifies the treatment of patients with neuroblastoma on the basis of a combination of biomarkers that include age and tumor histology classified by age-linked International Neuroblastoma Pathology Classification (INPC) criteria. By definition, this leads to a duplication of the prognostic contribution of age. The individual histologic features underlying the INPC have prognostic strength and are incorporated in the International Neuroblastoma Risk Group classification schema. Here, we analyzed data in the International Neuroblastoma Risk Group Data Commons to validate the prognostic strength of the underlying INPC criteria and to determine whether a risk classification devoid of the confounding of age and INPC criteria will identify new prognostic subgroups.

PATIENTS AND METHODS

Event-free survival of patients diagnosed between 1990 and 2002 (cohort 1; n = 10,104) and between 2003 and 2016 (cohort 2; n = 8,761) was analyzed. Recursive partitioning with univariate Cox models of event-free survival ("survival tree regression") was performed using (1) individual INPC criteria (age at diagnosis, histologic category, mitosis-karyorrhexis index (MKI), grade of differentiation) and (2) factors in (1) plus other COG-risk biomarkers (International Neuroblastoma Staging System [INSS] stage, MYCN status, ploidy).

RESULTS

The independent prognostic ability of age, histologic category, MKI, and grade were validated. Four histologic prognostic groups were identified (< 18 months with low v high MKI, and ≥ 18 months with differentiating v undifferentiated/poorly differentiating tumors). Compared with survival trees generated with established COG risk criteria, an additional prognostic subgroup was identified and validated when individual histologic features were analyzed in lieu of INPC.

CONCLUSION

Replacing INPC with individual histologic features in the COG risk classification will eliminate confounding, facilitate international harmonization of risk classification, and provide a schema for more precise prognostication and refined therapeutic approaches.

5-Hydroxymethylcytosine Profiles in Circulating Cell-Free DNA Associate with Disease Burden in Children with Neuroblastoma

PURPOSE

5-Hydroxymethylcytosine (5-hmC) is an epigenetic marker of open chromatin and active gene expression. We profiled 5-hmC with Nano-hmC-Seal technology using 10 ng of plasma-derived cell-free DNA (cfDNA) in blood samples from patients with neuroblastoma to determine its utility as a biomarker.

EXPERIMENTAL DESIGN

For the Discovery cohort, 100 5-hmC profiles were generated from 34 well children and 32 patients (27 high-risk, 2 intermediate-risk, and 3 low-risk) at various time points during the course of their disease. An independent Validation cohort encompassed 5-hmC cfDNA profiles (n = 29) generated from 21 patients (20 high-risk and 1 intermediate-risk). Metastatic burden was classified as high, moderate, low, or none per Curie metaiodobenzylguanidine scores and percentage of tumor cells in bone marrow. Genes with differential 5-hmC levels between samples according to metastatic burden were identified using DESeq2.

RESULTS

Hierarchical clustering using 5-hmC levels of 347 genes identified from the Discovery cohort defined four clusters of samples that were confirmed in the Validation cohort and corresponded to high, high-moderate, moderate, and low/no metastatic burden. Samples from patients with increased metastatic burden had increased 5-hmC deposition on genes in neuronal stem cell maintenance and epigenetic regulatory pathways. Further, 5-hmC cfDNA profiles generated with 1,242 neuronal pathway genes were associated with subsequent relapse in the cluster of patients with predominantly low or no metastatic burden (sensitivity 65%, specificity 75.6%).

CONCLUSIONS

cfDNA 5-hmC profiles in children with neuroblastoma correlate with metastatic burden and warrants development as a biomarker of treatment response and outcome.

Pembrolizumab in paediatric patients with advanced melanoma or a PD-L1-positive, advanced, relapsed, or refractory solid tumour or lymphoma (KEYNOTE-051): interim analysis of an open-label, single-arm, phase 1-2 trial

BACKGROUND

Pembrolizumab is approved for the treatment of advanced cancer in adults; however, no information is available on safety and efficacy in paediatric patients. We aimed to establish the recommended phase 2 dose of pembrolizumab and its safety and antitumour activity in advanced paediatric cancer.

METHODS

KEYNOTE-051 is an ongoing phase 1-2 open-label trial. In this interim analysis, children aged 6 months to 17 years were recruited at 30 hospitals located in Australia, Brazil, Canada, France, Germany, Israel, Italy, South Korea, Sweden, the UK, and the USA. Patients with melanoma or a centrally confirmed, PD-L1-positive, relapsed or refractory solid tumour or lymphoma, and a Lansky Play/Karnofsky Performance status score of 50 or higher, received intravenous pembrolizumab at an initial dose of 2 mg/kg every 3 weeks. Pharmacokinetics and dose-limiting toxicities were used to establish the recommended phase 2 dose, and the safety and antitumour activity of this dose were assessed. Primary endpoints were determination of dose-limiting toxicities at the maximum administered dose, safety and tolerability, and the proportion of patients with objective response to pembrolizumab for each tumour type according to the Response Evaluation Criteria in Solid Tumours version 1.1 or the International Neuroblastoma Response Criteria. Safety and efficacy were assessed in all treated patients who received at least one dose of pembrolizumab. Separate reporting of the cohort of patients with relapsed or refractory classical Hodgkin lymphoma was a post-hoc decision. The data cutoff for this interim analysis was Sept 3, 2018. This trial is still enrolling patients and is registered with ClinicalTrials.gov, number NCT02332668.

FINDINGS

Of 863 patients screened between March 23, 2015, and Sept 3, 2018, 796 had tumours that were evaluable for PD-L1 expression (278 [35%] were PD-L1-positive); 155 eligible patients were enrolled and 154 had at least one dose of pembrolizumab. The median age of the enrolled patients was 13 years (IQR 8-15). Median follow-up was 8·6 months (IQR 2·5-16·4). No dose-limiting toxicities were reported in phase 1, and pembrolizumab plasma concentrations were consistent with those previously reported in adults; the recommended phase 2 dose was therefore established as 2 mg/kg every 3 weeks. Of the 154 patients treated, 69 (45%) experienced grade 3-5 adverse events, most commonly anaemia in 14 (9%) patients and decreased lymphocyte count in nine (6%) patients. 13 (8%) of the 154 patients had grade 3-5 treatment-related adverse events, most commonly decreased lymphocyte count in three (2%) patients and anaemia in two (1%) patients. 14 (9%) patients had serious treatment-related adverse events, most commonly pyrexia (four [3%]), and hypertension and pleural effusion (two [1%] each). Four patients (3%) discontinued treatment because of treatment-related adverse events, and two (1%) died (one due to pulmonary oedema and one due to pleural effusion and pneumonitis). Of 15 patients with relapsed or refractory Hodgkin lymphoma, two had complete and seven had partial responses; thus, nine patients achieved an objective response (60·0%; 95% CI 32·3-83·7). Of 136 patients with solid tumours and other lymphomas, eight had partial responses (two patients each with adrenocortical carcinoma and mesothelioma, and one patient each with malignant ganglioglioma, epithelioid sarcoma, lymphoepithelial carcinoma, and malignant rhabdoid tumour); the proportion of patients with an objective response was 5·9% (95% CI 2·6-11·3).

INTERPRETATION

Pembrolizumab was well tolerated and showed encouraging antitumour activity in paediatric patients with relapsed or refractory Hodgkin lymphoma, consistent with experience in adult patients. Pembrolizumab had low antitumour activity in the majority of paediatric tumour types, and responses were observed in only a few rare PD-L1-positive tumour types, suggesting that PD-L1 expression alone is not sufficient as a biomarker for the selection of paediatric patients who are likely to respond to PD-1 checkpoint inhibitors. Final results of KEYNOTE-051, expected by September, 2022, with the possibility for extension, will report further on the activity of pembrolizumab in Hodgkin lymphoma, microsatellite instability-high tumours, and melanoma.

FUNDING

Merck Sharp & Dohme, a subsidiary of Merck & Co.

Statistical Framework in Support of a Revised Children's Oncology Group Neuroblastoma Risk Classification System

PURPOSE

The International Neuroblastoma Risk Group (INRG) Staging System (INRGSS) was developed through international consensus to provide a presurgical staging system that uses clinical and imaging data at diagnosis. A revised Children's Oncology Group (COG) neuroblastoma (NB) risk classification system is needed to incorporate the INRGSS and within the context of modern therapy. Herein, we provide statistical support for the clinical validity of a revised COG risk classification system.

PATIENTS AND METHODS

Nine factors were tested for potential statistical and clinical significance in 4,569 patients diagnosed with NB who were enrolled in the COG biology/banking study ANBL00B1 (2006-2016). Recursive partitioning was performed to create a survival-tree regression (STR) analysis of event-free survival (EFS), generating a split by selecting the strongest prognostic factor among those that were statistically significant. The least absolute shrinkage and selection operator (LASSO) was applied to obtain the most parsimonious model for EFS. COG patients were risk classified using STR, LASSO, and per the 2009 INRG classification (generated using an STR analysis of INRG data). Results were descriptively compared among the three classification approaches.

RESULTS

The 3-year EFS and overall survival (± SE) were 72.9% ± 0.9% and 84.5% ± 0.7%, respectively (N = 4,569). In each approach, the most statistically and clinically significant factors were diagnostic category (eg, NB, ganglioneuroblastoma), INRGSS, MYCN status, International Neuroblastoma Pathology Classification, ploidy, and 1p/11q status. The results of the STR analysis were more concordant with those of the INRG classification system than with LASSO, although both methods showed moderate agreement with the INRG system.

CONCLUSION

These analyses provide a framework to develop a new COG risk classification incorporating the INRGSS. There is statistical evidence to support the clinical validity of each of the three classifications: STR, LASSO, and INRG.

Unrealistic parental expectations for cure in poor-prognosis childhood cancer

BACKGROUND

Many parents of children with advanced cancer pursue curative goals when cure is no longer possible. To the authors' knowledge, no pediatric studies to date have prospectively evaluated prognosis communication or influences on decision making in poor-prognosis childhood cancer.

METHODS

The authors conducted a prospective cohort study at 9 pediatric cancer centers that enrolled 95 parents of children with recurrent or refractory, high-risk neuroblastoma (63% of those who were approached), a condition for which cure rarely is achieved. Parents were surveyed regarding the child's likelihood of cure; their primary goal of care; the child's symptoms, suffering, and quality of life; and regret concerning the last treatment decision. Medical records identified care and treatment decisions.

RESULTS

Only 26% of parents recognized that the chance of cure was <25%. When asked to choose a single most important goal of care, approximately 72% chose cure, 10% chose longer life, and 18% chose quality of life. Parents were more likely to prioritize quality of life when they recognized the child's poor prognosis (P = .002). Approximately 41% of parents expressed regret about the most recent treatment decision. Parents were more likely to experience regret if the child had received higher intensity medical care (odds ratio [OR], 3.14; 95% CI, 1.31-7.51), experienced suffering with limited benefit from the most recent treatment (OR, 4.78; 95% CI, 1.16-19.72), or experienced suffering from symptoms (OR, 2.91; 95% CI, 1.18-7.16).

CONCLUSIONS

Parents of children with poor-prognosis cancer frequently make decisions based on unrealistic expectations. New strategies for effective prognosis communication are needed.

Maintaining Outstanding Outcomes Using Response- and Biology-Based Therapy for Intermediate-Risk Neuroblastoma: A Report From the Children's Oncology Group Study ANBL0531

PURPOSE

The primary objective of the Children's Oncology Group study ANBL0531 (ClinicalTrials.gov identifier: NCT00499616) was to reduce therapy for subsets of patients with intermediate-risk neuroblastoma using a biology- and response-based algorithm to assign treatment duration while maintaining a 3-year overall survival (OS) of 95% or more for the entire cohort.

PATIENTS AND METHODS

Children younger than age 12 years with intermediate-risk stage 2A/2B or stage 3 tumors with favorable histology; infants younger than age 365 days with stage 3, 4 or 4S disease; and toddlers from 365 to younger than 547 days with favorable histology, hyperdiploid stage 4, or unfavorable histology stage 3 tumors were eligible. Patients with MYCN-amplified tumors were excluded. Patients were assigned to initially receive two (group 2), four (group 3), or eight (group 4) cycles of chemotherapy with or without surgery on the basis of prognostic markers, including allelic status of chromosomes 1p and 11q; ultimate duration of therapy was determined by overall response.

RESULTS

Between 2007 and 2011, 404 evaluable patients were enrolled. Compared with legacy Children's Oncology Group studies, subsets of patients had a reduction in treatment. The 3-year event-free survival and OS rates were 83.2% (95% CI, 79.4% to 87.0%) and 94.9% (95% CI, 92.7% to 97.2%), respectively. Infants with stage 4 tumors with favorable biology (n = 61) had superior 3-year event-free survival compared with patients with one or more unfavorable biologic features (n = 47; 86.9% [95% CI, 78.3% to 95.4%] v 66.8% [95% CI, 53.1% to 80.6%]; P = .02), with a trend toward OS advantage (95.0% [95% CI, 89.5% to 100%] v 86.7% [95% CI, 76.6% to 96.7%], respectively; P = .08). OS for patients with localized disease was 100%.

CONCLUSION

Excellent survival was achieved with this treatment algorithm, with reduction of therapy for subsets of patients. More-effective treatment strategies still are needed for infants with unfavorable biology stage 4 disease.

Role of the extent of prophylactic regional lymph node radiotherapy on survival in high-risk neuroblastoma: A report from the COG A3973 study

PURPOSE

Neuroblastoma is the most common extracranial solid pediatric malignancy, with poor outcomes in high-risk disease. Standard treatment approaches employ an increasing array of aggressive multimodal therapies, of which local control with surgery and radiotherapy remains a backbone; however, the benefit of broad regional nodal irradiation remains controversial. We analyzed centrally reviewed radiation therapy data from patients enrolled on COG A3973 to evaluate the impact of primary site irradiation and the extent of regional nodal coverage stratified by extent of surgical resection.

METHODS

Three hundred thirty high-risk neuroblastoma patients with centrally reviewed radiotherapy plans were analyzed. Outcome was evaluated by the extent of nodal irradiation. For the 171 patients who also underwent surgery (centrally reviewed), outcome was likewise analyzed according to the extent of resection. Overall survival (OS), event-free survival (EFS), and cumulative incidence of local progression (CILP) were examined by Kaplan-Meier, log-rank test (EFS, OS), and Grey test (CILP).

RESULTS

The five-year CILP, EFS, and OS for all 330 patients receiving radiotherapy on A3973 were 8.5% ± 1.5%, 47.2% ± 3.0%, and 59.7% ± 3.0%, respectively. There were no significant differences in outcomes based on the extent of lymph node irradiation regardless of the degree of surgical resection (< 90% or ≥90%).

CONCLUSION

Although local control remains a significant component of treatment of high-risk neuroblastoma, our results suggest there is no benefit of extensive lymph node irradiation, irrespective of the extent of surgical resection preceding stem cell transplant.

5-Hydroxymethylcytosine Profiles Are Prognostic of Outcome in Neuroblastoma and Reveal Transcriptional Networks That Correlate With Tumor Phenotype

PURPOSE

Whole-genome profiles of the epigenetic modification 5-hydroxymethylcytosine (5-hmC) are robust diagnostic biomarkers in adult patients with cancer. We investigated if 5-hmC profiles would serve as novel prognostic markers in neuroblastoma, a clinically heterogeneous pediatric cancer. Because this DNA modification facilitates active gene expression, we hypothesized that 5-hmC profiles would identify transcriptomic networks driving the clinical behavior of neuroblastoma.

PATIENTS AND METHODS

Nano-hmC-Seal sequencing was performed on DNA from Discovery (n = 51), Validation (n = 38), and Children’s Oncology Group (n = 20) cohorts of neuroblastoma tumors. RNA was isolated from 48 tumors for RNA sequencing. Genes with differential 5-hmC or expression between clusters were identified using DESeq2. A 5-hmC model predicting outcome in high-risk patients was established using linear discriminant analysis.

RESULTS

Comparison of low- versus high-risk tumors in the Discovery cohort revealed 577 genes with differential 5-hmC. Hierarchical clustering of tumors from the Discovery and Validation cohorts using these genes identified two main clusters highly associated with established prognostic markers, clinical risk group, and outcome. Genes with increased 5-hmC and expression in the favorable cluster were enriched for pathways of neuronal differentiation and KRAS activation, whereas genes involved in inflammation and the PRC2 complex were identified in the unfavorable cluster. The linear discriminant analysis model trained on high-risk Discovery cohort tumors was prognostic of outcome when applied to high-risk tumors from the Validation and Children’s Oncology Group cohorts (hazard ratio, 3.8).

CONCLUSION

5-hmC profiles may be optimal DNA-based biomarkers in neuroblastoma. Analysis of transcriptional networks regulated by these epigenomic modifications may lead to a deeper understanding of drivers of neuroblastoma phenotype.

The challenge of defining "ultra-high-risk" neuroblastoma

Given the biological and clinical heterogeneity of neuroblastoma, risk stratification is vital to determining appropriate treatment. Historically, most patients with high-risk neuroblastoma (HR-NBL) have been treated uniformly without further stratification. Attempts have been made to identify factors that can be used to risk stratify these patients and to characterize an "ultra-high-risk" (UHR) subpopulation with particularly poor outcome. However, among published data, there is a lack of consensus in the definition of the UHR population and heterogeneity in the endpoints and statistical methods used. This review summarizes our current understanding of stratification of HR-NBL and discusses the complex issues in defining UHR neuroblastoma.

Maternal Embryonic Leucine Zipper Kinase (MELK), a Potential Therapeutic Target for Neuroblastoma

Maternal embryonic leucine zipper kinase (MELK) activates pathways that mediate aggressive tumor growth and therapy resistance in many types of adult cancers. Pharmacologic and genomic inhibition of MELK impairs tumor growth and increases sensitivity to radiation and chemotherapy. On the basis of these promising preclinical studies, early-phase adult clinical trials testing the MELK inhibitor OTS167 are ongoing. To investigate whether MELK is also a therapeutic target in neuroblastoma, we analyzed MELK expression in primary tumors and cell lines, and examined the effects of OTS167 on neuroblastoma growth. In primary tumors, high levels of MELK were associated with advanced stage disease and inferior survival. Higher levels of MELK were also detected in tumorigenic versus nontumorigenic neuroblastoma cell lines, and cells with higher levels of MELK expression were more sensitive to OTS167 than low-MELK expressing cells. OTS167 suppressed the growth of neuroblastoma xenografts, and in a preclinical model of minimal residual disease, survival was prolonged with MELK inhibition. OTS167 treatment downregulated MELK and its target enhancer of zeste homolog 2 (EZH2), a component of the polycomb repressive complex 2 (PRC2) that is known to modulate the DNA damage response. We also show that OTS167 reduced the formation of collapsed replication forks induced by camptothecin or radiation. Taken together, our results indicate that MELK indirectly mediates efficient processing of replication-associated DNA lesions in neuroblastoma, and that OTS167 sensitizes cells to DNA-damaging agents by abrogating this process. Further studies evaluating the activity of combination treatment regimens with OTS167 in neuroblastoma are warranted.

Phase II Trial of Alisertib in Combination with Irinotecan and Temozolomide for Patients with Relapsed or Refractory Neuroblastoma

PURPOSE

In phase I testing, alisertib tablets with irinotecan and temozolomide showed significant antitumor activity in patients with neuroblastoma. This study sought to confirm activity of this regimen; evaluate an alisertib oral solution; and evaluate biomarkers of clinical outcomes.

PATIENTS AND METHODS

We conducted a two-stage phase II trial of alisertib tablets (60 mg/m2/dose × 7 days), irinotecan (50 mg/m2/dose i.v. × 5 days), and temozolomide (100 mg/m2/dose orally × 5 days) in patients with relapsed or refractory neuroblastoma. The primary endpoint was best objective response. A separate cohort was treated with alisertib at 45 mg/m2 using oral solution instead of tablets. Exploratory analyses sought to identify predictors of toxicity, response, and progression-free survival (PFS) using pooled data from phase I, phase II, and oral solution cohorts.

RESULTS

Twenty and 12 eligible patients were treated in the phase II and oral solution cohorts, respectively. Hematologic toxicities were the most common adverse events. In phase II, partial responses were observed in 19 evaluable patients (21%). The estimated PFS at 1 year was 34%. In the oral solution cohort, 3 patients (25%) had first cycle dose-limiting toxicity (DLT). Alisertib oral solution at 45 mg/m2 had significantly higher median C max and exposure compared with tablets at 60 mg/m2. Higher alisertib trough concentration was associated with first cycle DLT, whereas MYCN amplification was associated with inferior PFS.

CONCLUSIONS

This combination shows antitumor activity, particularly in patients with MYCN nonamplified tumors. Data on an alisertib oral solution expand the population able to be treated with this agent.

Computer-assisted Curie scoring for metaiodobenzylguanidine (MIBG) scans in patients with neuroblastoma

BACKGROUND

Radiolabeled metaiodobenzylguanidine (MIBG) is sensitive and specific for detecting neuroblastoma. The extent of MIBG-avid disease is assessed using Curie scores. Although Curie scoring is prognostic in patients with high-risk neuroblastoma, there is no standardized method to assess the response of specific sites of disease over time. The goal of this study was to develop approaches for Curie scoring to facilitate the calculation of scores and comparison of specific sites on serial scans.

PROCEDURE

We designed three semiautomated methods for determining Curie scores, each with increasing degrees of computer assistance. Method A was based on visual assessment and tallying of MIBG-avid lesions. For method B, scores were tabulated from a schematic that associated anatomic regions to MIBG-positive lesions. For method C, an anatomic mesh was used to mark MIBG-positive lesions with automatic assignment and tallying of scores. Five imaging physicians experienced in MIBG interpretation scored 38 scans using each method, and the feasibility and utility of the methods were assessed using surveys.

RESULTS

There was good reliability between methods and observers. The user-interface methods required 57 to 110 seconds longer than the visual method. Imaging physicians indicated that it was useful that methods B and C enabled tracking of lesions. Imaging physicians preferred method B to method C because of its efficiency.

CONCLUSIONS

We demonstrate the feasibility of semiautomated approaches for Curie score calculation. Although more time was needed for strategies B and C, the ability to track and document individual MIBG-positive lesions over time is a strength of these methods.

Defining Risk Factors for Chemotherapeutic Intervention in Infants With Stage 4S Neuroblastoma: A Report From Children's Oncology Group Study ANBL0531

PURPOSE

Infants with stage 4S neuroblastoma usually have favorable outcomes with observation or minimal chemotherapy. However, young infants with symptoms secondary to massive hepatomegaly or with unfavorable tumor biology are at high risk of death. Our aim was to improve outcomes for patients with symptomatic and/or unfavorable biology 4S neuroblastoma with a uniform treatment approach using a biology- and response-based algorithm.

PATIENTS AND METHODS

The subset of patients with 4S disease with MYCN-not amplified tumors with impaired or impending organ dysfunction, or with unfavorable histology and/or diploid DNA index, were eligible. Patients were assigned to receive two, four, or eight cycles of chemotherapy on the basis of histology, diploid DNA index, chromosome arm 1p or 11q loss of heterozygosity (LOH) status, and symptoms.

RESULTS

Forty-nine eligible patients were enrolled: 41 were symptomatic and 28 had unfavorable biology. Seventeen patients (symptomatic, favorable biology) were assigned two cycles, 21 patients (any unfavorable biologic feature without 1p or 11q LOH) were assigned four cycles, and 11 patients (unfavorable biology including 1p and/or 11q LOH [n = 7] or symptomatic with unknown biology [n = 4]), were assigned eight cycles. The 3-year overall survival was 81.4% ± 5.8%. Eight of nine deaths were in patients younger than 2 months of age at diagnosis (median, 9 days [range, 1 to 68 days]): five acute deaths were a result of hepatomegaly and associated toxicities; two were a result of late relapse in patients with unfavorable biology; and two were a result of treatment complications. No deaths occurred after protocol-mandated pre-emptive treatment of infants younger than 2 months with hepatomegaly, regardless of symptoms. A new scoring algorithm for emergent chemotherapy in patients with 4S disease was developed on the basis of this experience.

CONCLUSION

The outcome for 4S neuroblastoma can be improved with pre-emptive chemotherapy for evolving hepatomegaly or other baseline comorbidities in infants younger than 2 months of age.

Abstract PR16: Whole-genome analysis of the epigenetic mark 5-hydroxymethylcytosine reveals differential profiles in low- , intermediate- , and high-risk neuroblastomas

Background: Neuroblastoma is characterized by a paucity of somatic mutations, suggesting epigenetic modifications likely play a key role in regulating gene expression driving tumor phenotype. 5-Methylcytosine is an epigenetic DNA modification that decreases transcription and silences several neuroblastoma tumor suppressors. New technology allows evaluation of genome-wide 5-hydroxymethylcytosine (5hmC), a marker of activated transcription. We hypothesize that linking whole-genome 5hmC profiling with tumor transcriptomes will identify epigenetically regulated oncogenic drivers of neuroblastoma phenotype, and may ultimately lead to the discovery of new therapeutic targets.

Methods: 5hmC was quantified by nano-hmC-Seal-Seq using 100ng of DNA from locally banked diagnostic neuroblastoma tumors. 50bp sequencing reads underwent quality control with Trimmomatic and were aligned to hg19 with Bowtie2. Homer software quantified peaks by genomic feature. Raw fragment counts were called using featureCounts and normalized with DeSeq2, which also identified genes with differential 5hmC. Gene Ontology (GO) pathway analysis of differential 5hmC genes was performed with the goseq package in R. 5hmC patterns were compared to gene expression in the E-MTAB-1781 cohort from ArrayExpress.

Results: 53 patients (27 low- , 12 intermediate- , and 14 high-risk) were evaluated. Low-risk tumors had increased 5hmC peaks per sample compared to high-risk tumors (140,062 vs. 79,727, p=0.014). Peaks were predominantly intronic in all risk groups. Global 5hmC profiles distinguished four groups of patients, low- , intermediate- , MYCN-amplified high-risk, and MYCN-nonamplified high-risk. After controlling for sex, batch, and MYCN-amplification status, 578 genes had differential 5hmC peaks between low- and high-risk tumors (padjusted&lt;0.05). Genes with increased 5hmC in low-risk tumors were enriched for GO pathways of neuronal differentiation, including cell morphogenesis involved in differentiation (p=9.72e-6), neuron projection morphogenesis (p=1.15e-5), and cell projection morphogenesis (p=1.89e-5). High-risk tumor genes were enriched genes in embryonal GO pathways, including anatomic structure formation involved in morphogenesis (p=9.72e-6), chordate embryonic development (p=1.15e-5), and embryo development (p=1.89e-5). Of the 578 genes with differential 5hmC, 149 were differentially expressed genes between 367 low-risk and 254 high-risk patients (poverlap&lt;1e-5) and 158 were significantly differentially expressed between survivors and deceased patients (poverlap&lt;1e-5). These 149 and 158 genes were enriched for neuronal pathways. MYCN-amplified high-risk tumors have divergent 5hmC patterns with MYCN-nonamplified tumors having decreased 5hmC on chromosomes 1p, 4q, 12q, and 17p.

Conclusions: We demonstrate that 5hmC profiles are able to different neuroblastoma risk groups and highlight cellular mechanisms important for tumor growth. These results suggest that neuroblastoma epigenetics play a contributory role to gene expression and tumor phenotype. Further analysis is ongoing to identify biologic drivers of these differences, which may be used to identify new therapeutic approaches.

Citation Format: Mark A. Applebaum, Erin Barr, Ji Nie, Sakshi Uppal, Wei Zhang, Alexandre Chlenski, Helen Salwen, Marija Dobratic, Barbara Stranger, Lucy Godley, Chuan He, Susan L. Cohn. Whole-genome analysis of the epigenetic mark 5-hydroxymethylcytosine reveals differential profiles in low- , intermediate- , and high-risk neuroblastomas [abstract]. In: Proceedings of the AACR Special Conference: Pediatric Cancer Research: From Basic Science to the Clinic; 2017 Dec 3-6; Atlanta, Georgia. Philadelphia (PA): AACR; Cancer Res 2018;78(19 Suppl):Abstract nr PR16.

Integrative genomics reveals hypoxia inducible genes that are associated with a poor prognosis in neuroblastoma patients

Neuroblastoma is notable for its broad spectrum of clinical behavior ranging from spontaneous regression to rapidly progressive disease. Hypoxia is well known to confer a more aggressive phenotype in neuroblastoma. We analyzed transcriptome data from diagnostic neuroblastoma tumors and hypoxic neuroblastoma cell lines to identify genes whose expression levels correlate with poor patient outcome and are involved in the hypoxia response. By integrating a diverse set of transcriptome datasets, including those from neuroblastoma patients and neuroblastoma derived cell lines, we identified nine genes (SLCO4A1, ENO1, HK2, PGK1, MTFP1, HILPDA, VKORC1, TPI1, and HIST1H1C) that are up-regulated in hypoxia and whose expression levels are correlated with poor patient outcome in three independent neuroblastoma cohorts. Analysis of 5-hydroxymethylcytosine and ENCODE data indicate that at least five of these nine genes have an increase in 5-hydroxymethylcytosine and a more open chromatin structure in hypoxia versus normoxia and are putative targets of hypoxia inducible factor (HIF) as they contain HIF binding sites in their regulatory regions. Four of these genes are key components of the glycolytic pathway and another three are directly involved in cellular metabolism. We experimentally validated our computational findings demonstrating that seven of the nine genes are significantly up-regulated in response to hypoxia in the four neuroblastoma cell lines tested. This compact and robustly validated group of genes, is associated with the hypoxia response in aggressive neuroblastoma and may represent a novel target for biomarker and therapeutic development.