A Bridge Over Troubled Water - Extending Induction for High-risk Neuroblastoma Patients With Poor End of Induction Response

Precis:

A subset of high-risk neuroblastoma patients who have a poor end of Induction response may benefit from receiving chemoimmunotherapy, as bridge therapy, prior to Consolidation. Prospective studies evaluating this therapeutic approach and improved identification of biomarkers of response are warranted.

While intensification of therapy including Induction,¹ Consolidation,^2,3 and post-Consolidation⁴ have advanced the cure rate of patients with high-risk neuroblastoma, approximately half of these patients succumb to disease.³ Despite these therapeutic advances, patients who have a poor End of Induction (EOI) response still have a poor prognosis. Pinto and colleagues performed a retrospective analysis evaluating over 1,200 patients from four consecutive Children’s Oncology Group (COG) high-risk neuroblastoma trials (A3973, ANBL02P1, ANBL0532 and ANBL12P1). In this analysis, an EOI response less than a partial response (PR) was associated with a significantly lower 3-year event free survival (EFS; 21% versus 54%; p<0.0001) and 3-year overall survival (OS) 46% versus 73% (p<0.0001).⁵ While these data provided valuable insight about the prognosis of patients with a poor EOI response, the field has been plagued by a paucity of data that describe the clinical courses of the EOI poor responders, including subsequent administered treatments. Further, little is known about patients enrolled on clinical trials who discontinue protocol therapy due to poor EOI response and/or receive therapy outside of a clinical trial.

Understanding the clinical course of patients with a poor EOI response is especially critical in the modern era of targeted therapy^6,7 and successful chemoimmunotherapy.^{1,8,9 131}I-metaiodobenzylguanidine (MIBG) with or without chemotherapy has demonstrated an objective response of approximately 15–30% in patients with recurrent or refractory (RR) disease.⁶ More recently, chemoimmunotherapy using a combination of an anti-GD2 monoclonal antibody and different chemoimmunotherapeutic regimens in the RR neuroblastoma population, has demonstrated the most robust response rates to date.^8,9 In the largest study, ANBL1221, patients with RR neuroblastoma randomized to treatment with dinutuximab, irinotecan and temozolomide (DIT) had an objective response rate (≥ PR) of 53%.¹⁰ The trial expanded accrual in order to better evaluate the response rate and toxicity profile of this combination. In the combined chemoimmunotherapy cohorts, a 41.5% response rate was observed, including a 32.3% response rate in patients with refractory or progressive disease.⁹ The results of this trial led to a shift in clinical practice to administer DIT as a standard of care for patients with refractory disease with the goal of “getting them back on track” and proceeding to tandem ASCT, albeit without data demonstrating this improves outcomes.

In the article, “Efficacy of Post-Induction Therapy for High-risk Neuroblastoma Patients with End-Induction Residual Disease”, Desai and colleagues retrospectively assessed the clinical course and outcome of patients treated at 6 large pediatric centers who had an EOI response of ≤ PR. Specifically, they evaluated patients with EOI residual disease who received “bridge therapy” prior to receiving Consolidation (high-dose chemotherapy followed by autologous stem cell transplant [ASCT]) and compared them to a group of patients who did not receive bridge therapy. Bridge therapy most often included chemoimmunotherapy or MIBG therapy.

For the Desai analysis, the patients were divided into three cohorts: 1) no bridge therapy prior to ASCT, 2) bridge therapy prior to ASCT and 3) bridge therapy without ASCT. Not surprisingly, the groups were skewed with a higher percentage of patients who received bridge therapy having had a worse EOI response compared to Group 1 patients who did not receive bridge therapy. Interestingly, while the Pinto et al. data would suggest that cohorts 2 and 3 should therefore have a worse outcome due to the disproportionate number of patients with EOI response < PR, Desai demonstrated that patients in Cohort 2, who received bridge therapy prior to Consolidation, had a similar outcome to those in Cohort 1. Furthermore, those patients who were able to achieve a metastatic complete response (CR) had a statistically significant improvement in outcome over those with EOI metastatic SD who did not receive bridge therapy. This suggests that a subset of patients may benefit from additional therapy prior to ASCT with the goal to achieve improved metastatic response prior to Consolidation.

It is difficult to draw significant conclusions from the patients included in Cohort 3, who received bridge therapy but did not receive ASCT. As the ANBL0532 study demonstrated, some patients with high-risk neuroblastoma who had an EOI response ≤ PR ultimately had improved outcome related to ASCT.³ Thus, it is not entirely surprising that the patients included in Cohort 3, who did not receive ASCT, had inferior outcome compared to Cohorts 1 and 2. However, a small subset of patients treated in Cohort 3, who achieved a metastatic CR following bridge therapy, remain alive as of last follow-up. This highlights the necessity to further evaluate data from multi-institutional and cooperative group trials to better identify which patients with high-risk disease may achieve excellent outcome without the need for ASCT. This is a critical question that the field will need to address in the future.¹¹ Further, it highlights the need for identification of better biomarkers of response in neuroblastoma assessments.

Prior studies from the North American and European cooperative groups have analyzed the prognostic value of EOI Curie scores and concluded that patients with EOI Curie scores > 2 and SIOPEN MIBG skeletal scores > 3 have inferior outcomes.^12,13 However, nearly all of the patients included in these prior analyses were treated without post-Consolidation immunotherapy. Thus, the optimal EOI Curie cut point, in the context of modern high-risk neuroblastoma therapy, remains unknown. Desai and colleagues did not provide data on EOI Curie scores; rather, they analyzed patients using the International Neuroblastoma Response Criteria (INRC).¹⁴ While their data are compelling, we caution clinicians against continuing to treat patients until they achieve a metastatic CR without additional prospective studies to validate this approach. The Desai dataset included very few patients (7%) with EOI SD, and so it is not clear that the “metastatic CR bar” can be applied to all patients with EOI < PR. This approach may delay Consolidation and/or could lead to over-treatment and increased toxicity without additional improvement in outcomes. Future prospective studies analyzing overall INRC response, including individual INRC response components (metastatic soft tissue and bone, primary site, and bone marrow response), and EOI Curie scores in a modern cohort of high-risk neuroblastoma patients are needed.

Despite extensive efforts to identify biomarkers of response to guide therapeutic changes, the neuroblastoma community still lacks a robust method to identify which patients are likely to fail treatment, and importantly which can potentially be salvaged. Known variables including age, stage, MYCN amplification and segmental chromosomal aberrations are prognostic and are included in risk classification.¹⁵ However, unlike the evaluation of minimal residual disease (MRD) in the treatment of acute lymphoblastic leukemia,^16,17 less is known about molecular biomarkers of neuroblastoma response to guide real time treatment modification. While preliminary data of candidate biomarkers such as telomere maintenance mechanisms, ctDNA, and circulating GD2 suggest that these factors are important, further study of these biomarkers in the context of modern high-risk neuroblastoma therapy is needed before they can be incorporated.^18–20 Future studies should evaluate whether these biomarkers can be utilized as stand-alone predictors of outcome and/or used in combination with other clinical factors and response.

Desai and colleagues provide strong retrospective data that supports the need to evaluate bridge therapy, using chemoimmunotherapy, in future studies for patients who experience a poor EOI response to standard Induction chemotherapy. These studies should be designed in a way that keeps patients who receive bridge therapy on study. Further, prospective studies should implement detailed data collection tools to better evaluate patients who come off protocol therapy so that outcomes for this patient population can be accurately described in the future.