Improving Outcomes in Children With High-Risk Neuroblastoma: The Role of Randomized Trials

Outcomes for children with newly diagnosed high-risk neuroblastoma have improved significantly over the past 20 years, but despite intensive multimodality therapy, survival rates remain just over 50%.¹ Modern-era therapy for high-risk patients with this embryonal malignancy in higher-income countries is composed of induction therapy that includes multiagent chemotherapy and surgery, consolidation therapy consisting of high-dose chemotherapy with autologous stem-cell rescue and external beam radiotherapy, and postconsolidation immunotherapy to address minimal residual disease.² Neuroblastoma predominantly affects young children, and late effects of therapy place a heavy burden on those who survive.

Investigators around the world have long been intrigued by the challenges posed by this disease, and international collaborative efforts have resulted in key insights into the biologic basis of neuroblastoma and in major advancements in its treatment. In the article that accompanies this editorial, Garaventa et al³ describe the results of the HR-NBL1.5 trial conducted by the International Society of Paediatric Oncology European Neuroblastoma Group (SIOPEN), a cooperative group that is committed to conducting randomized trials to address important questions that affect the care of patients with neuroblastoma. This large, multicenter study was conducted over nearly 16 years in total and included five randomizations that have addressed important questions related to all three phases of high-risk neuroblastoma therapy (induction, consolidation, and postconsolidation).

The current manuscript addresses a question regarding induction therapy, the phase of treatment designed to maximally reduce the burden of disease before subsequent therapy. In previous cooperative group studies, approximately 7%-15% of patients experienced early disease progression,^1,4,5 highlighting the importance of identifying the most effective initial treatment for high-risk patients. Furthermore, it has been shown that subsequent survival outcomes are superior in patients who experience a partial response or better during induction therapy compared with those who have stable or progressive disease initially.⁶ In the HR-NBL1.5 trial, the European regimen associated with the best reported end-induction outcomes (known as rapid cisplatin, vincristine, carboplatin, etoposide, and cyclophosphamide [COJEC]⁴) was compared with the North American regimen associated with the best reported end-induction outcomes (the N5 regimen developed by investigators at Memorial Sloan Kettering Cancer Center, or MSKCC N5).⁷ The response rate associated with the MSKCC N5 regimen was initially reported in the context of a single-arm trial conducted at a single North American center and was not replicated in other contexts.^8-10 Therefore, one might have argued that the preferred comparator would have been the induction regimen studied in multicenter cooperative group trials conducted by the Children's Oncology Group (COG) or another large consortium, given the greater expected generalizability of the results from multicenter trials compared with single-center studies. Nonetheless, the more important decision made by the SIOPEN team was the decision to commit to conducting a randomized clinical trial (RCT).

Although for decades RCTs have been widely acknowledged to be the gold standard in clinical trial design, in recent years, the value of RCTs has been called into question, particularly in the context of rare diseases. In the fast-moving world of modern medicine, the time it takes to perform an RCT runs counter to the desire for a rapid readout, particularly when new agents are being studied. Questions arise as to whether the results of a pediatric RCT that requires 5 years to accrue in a rare disease will still be relevant when the trial is completed. Agents may or may not have been found to be effective in adult trials by the time a pediatric RCT is completed and, therefore, the agent may or may not be readily accessible for the pediatric oncology population. In addition, by their very nature, RCTs inform our thinking about the potential benefits of a given regimen in a large group of patients who meet trial eligibility criteria. However, RCTs may or may not facilitate individualized decision making by clinicians and patients or parents if there are aspects of a particular patient's disease that are unusual, such as a specific molecular aberration that could render results that are applicable to the larger group less relevant to a specific patient at hand. In addition, clinical trialists must remain cognizant of concerns related to equity and inclusion, understanding that groups of patients may be under-represented in oncology trials unless efforts are made to ensure that trial populations truly reflect the population of patients affected by a given disease.

With these concerns in mind, is there still a role for RCTs in the setting of high-risk neuroblastoma clinical studies? The results of the SIOPEN trial tell us that the answer is yes. A total of 941 patients were enrolled on the HR-NBL1.5 study, a testament to the investment that SIOPEN was willing to make to definitively identify the induction regimen that would be considered standard in their future trials. The difference in duration of induction for patients assigned to rapid COJEC and MSKCC N5 was relatively small, and the proportions of patients in each arm who had objective responses and metastatic complete responses were similar, as were the proportions of patients who were able to proceed on to further therapy on the study. However, it appears that the burden of therapy-related toxicity was mitigated to some extent with rapid COJEC compared with MSKCC-N5, an important consideration for young children who are receiving very lengthy and intensive therapy for this disease. Patients were treated at 160 centers in more than 25 countries and, therefore, the diversity of participating sites is a real strength, particularly since toxicity was ultimately a very important consideration.

Although acute toxicities proved crucial in selecting the SIOPEN standard regimen based on data from this study, it is well known that patients with high-risk neuroblastoma suffer from a tremendous burden of late effects, including hearing loss, short stature, and risk of second malignancy.¹¹ Patients on the MSKCC-N5 arm of the trial had nominally higher rates of any grade ototoxicity and of grade 3-4 ototoxicity during the period of follow-up described in this manuscript. It will be important to follow patients on this trial into the future to understand whether other differences in late effects associated with these two induction regimens emerge over time. Large cooperative group trials such as HR-NBL1.5 have the potential to improve our understanding of less common but clinically important late effects that may be missed in smaller studies. However, this advantage of large trials can only be realized if there are mechanisms in place to capture long-term toxicities resulting from therapy. These mechanisms may be embedded within the trial, or trial participants may be offered participation in follow-on observational studies specifically designed to evaluate a wide range of potential late effects. For example, the ongoing COG Late Effects After High-Risk Neuroblastoma (LEAHRN) study is addressing this issue in survivors who were enrolled on high-risk neuroblastoma trials.¹¹

As RCTs are conducted to identify the regimens associated with better survival and toxicity outcomes between groups of patients treated on high-risk neuroblastoma trials, it is also important to identify the basis for differential response to therapy within groups of patients on such studies. In addition, it is critical that we use data derived from RCTs to develop predictors of differential response. Development of robust predictive biomarkers could be expected to allow for more precisely tailored treatment, thereby reserving more toxic components of therapy for patients who may derive the most benefit from them. The coprimary end point of metastatic complete response in the SIOPEN trial highlights current limitations on our ability to predict chemotherapy response in this disease. The SIOPEN group reported a significant association between tumor MYCN amplification and favorable chemotherapy response, validating the same observation from previous groups.^6,12 As the genomics of neuroblastoma have become increasingly well understood, several molecular features, including ALK aberrations and presence of telomere maintenance mechanisms, have been associated with inferior overall and/or event-free survival.^13-15 The impact of these additional biomarkers in predicting chemotherapy responsiveness remains largely unknown. The use of clinically annotated biospecimens from large cooperative group trials, such as HR-NBL1.5/SIOPEN, has the potential to provide important insights that may enable us to refine our initial treatment strategies for patients whose tumors harbor specific aberrations.

Although there is a great deal to be learned from RCTs such as HR-NBL1.5, the results of the trial reinforce the sobering reality that we have likely reached a therapeutic plateau in what can be achieved with conventional chemotherapy and primary site resection during initial induction therapy. Most modern cooperative group trials have yielded end-induction response rates of approximately 70%-80%, regardless of number of cycles or specific conventional chemotherapy agents used. In this context, new high-risk neuroblastoma trials should rigorously evaluate induction strategies that incorporate promising targeted therapies that have shown activity in the relapsed setting. For example, the ongoing COG ANBL1531 trial for children with newly diagnosed high-risk neuroblastoma includes a randomized component that evaluates ¹³¹I-metaiodobenzylguanidine during induction therapy for patients with metaiodobenzylguanidine-avid tumors that lack a detectable ALK aberration. Patients with tumors with an ALK mutation or ALK amplification are nonrandomly assigned to receive an ALK inhibitor along with conventional induction therapy. Likewise, the SIOPEN HR-NBL2 trial will evaluate the addition of an ALK inhibitor to rapid COJEC induction therapy for patients with tumors with an ALK aberration. These large, international trials will provide an opportunity to test the hypothesis that the early introduction of targeted therapies may help overcome the therapeutic plateau highlighted by the SIOPEN HR-NBL1.5 trial and result in better outcomes for children with high-risk disease.

AUTHOR CONTRIBUTIONS

Conception and design: All authors

Data analysis and interpretation: All authors

Manuscript writing: All authors

Final approval of manuscript: All authors

Accountable for all aspects of the work: All authors

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

Improving Outcomes in Children With High-Risk Neuroblastoma: The Role of Randomized Trials

The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated unless otherwise noted. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO’s conflict of interest policy, please refer to www.asco.org/rwc or ascopubs.org/jco/authors/author-center.

Open Payments is a public database containing information reported by companies about payments made to US-licensed physicians (Open Payments).