What Is Next in Pediatric B-Cell Precursor Acute Lymphoblastic Leukemia

Abstract

Cure rates now exceed 90% in many contemporary trials for children with B-cell acute lymphoblastic leukemia (B-ALL). However, treatment remains suboptimal, and therapy is toxic for all patients. New treatment options potentially offer the chance to reduce both treatment resistance and toxicity. Here, we review recent advances in ALL diagnostics, chemotherapy, and immunotherapy. In addition to describing recently published results, we also attempt to project the impact of these new developments into the future to imagine what B-ALL therapy may look like in the next few years.

1. Introduction

Acute lymphoblastic leukemia (ALL) is the most common cancer in children, with more than 50,000 children newly diagnosed each year globally [1]. In children with ALL, approximately 75% to 88% present with B-cell precursor immunophenotype disease, depending on their ancestry [2]. Serial clinical trials have improved overall survival (OS) in high resource settings to more than 85%, and many trial groups are now reporting OS rates over 90% at 5 years [3,4,5,6,7]. However, survival in resource-limited settings remains below 70%, and, due to a lack of disease detection and documentation in such areas, survival is likely much lower than these data suggest [1,5]. Moreover, specific molecular subtypes with prognostic significance are associated with genetic ancestry, accounting for some of the racial and ethnic disparities in outcomes [2]. Furthermore, for those patients with relapsed disease, outcomes are much poorer across treatment settings [8,9,10].

As we consider the future of therapy for B-ALL, we recognize that such a future will proceed along multiple paths simultaneously, including: (i) the use of novel agents to improve survival and reduce toxicity in patients treated in high-resource settings, (ii) the application of molecular subtype-driven treatment individualization to improve efficacy and to eliminate racial and ethnic gaps in outcomes, and (iii) the expansion of access to additional effective therapies in middle- and low-resource settings. Additional considerations will include the timing of novel therapies in the disease course, in both the up-front and relapsed patient populations; replacement of or reduction in traditional chemotherapy agents; advancements in the availability and quality of diagnostic modalities; and further development of targeted therapies for specific disease subtypes.

This review will focus on emerging classifications and treatment strategies that may play a larger role in the treatment of children with B-ALL in high-resource settings in the coming years. We will divide the discussion of therapeutic strategies into two categories, reflecting the intertwining modalities driving modern therapy for ALL: chemotherapy, including both conventional agents and new antigen-directed chemo-immunotherapy drugs; and cellular therapy, including allogeneic hematopoietic cell transplantation (HCT) and chimeric antigen receptor (CAR) T-cell therapy.

2. Improvements to Risk Stratification and Biological Classification

Despite differences in several important components of therapy, most trial groups are now reporting event-free survival (EFS) of over 80% [5]. Recent up-front treatment studies in pediatric B-ALL trials have evaluated the optimal steroid selection and schedule during induction [6,11,12], the role of high-dose methotrexate across different disease subsets [6,13,14], and modifications to risk stratification based on either minimal residual disease (MRD) or genomics [15]. Several additional studies have evaluated modifications in asparaginase formulation or its schedule [13,16,17,18,19,20,21,22]. While interactions between treatment backbone, risk stratification, and study intervention preclude the identification of a single optimal regimen for all patients, recent data have continued to refine the patient populations which are appropriate for therapy intensification or deintensification.

Looking forward, we anticipate that ongoing improvements in risk stratification will provide opportunities for further treatment modulation of initial therapy to reduce both acute and long-term toxicity for lower-risk patient subsets. Although it provides valuable prognostic information, the National Cancer Institute (NCI) risk stratification [23] for B-ALL alone is insufficient for optimal treatment intensification/deintensification [24,25]. The identification of low-level persistent disease following morphologic remission has historically utilized either flow cytometry (FC) or patient-specific T-cell receptor (TCR) or immunoglobulin heavy chain (IgH) polymerase chain reaction (PCR) testing as MRD methods. On the other hand, it has also relied on the persistence of molecular aberrancy in chromosomal analysis, florescence in situ hybridization, or detection of persistent gene fusions through PCR. These modalities dramatically improve risk stratification compared to morphology alone [16,26,27,28,29]. High-throughput sequencing (HTS, sometimes called next-generation sequencing)-based MRD testing provides the ability to detect residual leukemic clones at 10- to 100-fold lower levels than PCR or FC [30,31]. Importantly, HTS also appears to identify patients who appear to be MRD negative by conventional FC or PCR assays, but who have residual leukemia [32,33]. The HTS-based MRD test is helpful in identifying low-risk patients who have exceptionally favorable outcomes and those who are likely to relapse after allogeneic transplantation or CAR T-cell therapy [29,30,31,32,33,34]. Moreover, HTS-based MRD can also identify subgroups of patients who lack immunoglobulin gene rearrangements at diagnosis and have poor outcomes [30].

Ongoing efforts to integrate our evolving understanding of B-ALL genomics will further change the way we think about risk and treatment. Prior work has demonstrated that the prognostic impact of low-level MRD in patients with hyperdiploid ALL is lower than similarly low levels of MRD in either patients with ETV6∷RUNX1 ALL or other patients with B-ALL. This is likely related to the sensitivity of hyperdiploid ALL to antimetabolite therapies, which are not included in induction [35]. Indeed, patients with hyperdiploid ALL who fail to achieve complete remission after conventional induction therapy have a high salvage rate without the need of allogeneic transplantation, presumably because many of them are highly sensitive to the high-dose methotrexate and mercaptopurine that were used for post-remission therapy [36]. Additional work has integrated newly identified genomic subsets of ALL to further identify patients who perform poorly despite their apparent rapid responses to chemotherapy [37,38]. As comprehensive genomic analyses become more common, our understanding of the interactions between genomic lesions will expand, identifying combinations of lesions with prognostic and therapeutic significance. Such work has already begun, and we believe that progress in this area will continue to accelerate [39]. We anticipate that future treatments will intensify therapy for these patients despite the MRD response, mirroring strategies which are already being employed for patients with KMT2A-rearranged and hypodiploid ALL in many groups. Finally, we expect that newly identified ALL subsets will further offer opportunities for targeted additions or subtractions of therapy for these patients. In vitro drug sensitivity testing, integrated with genomic information, will provide opportunities to identify subtype-specific vulnerabilities or resistance patterns in order to enable such therapeutic modifications [40,41,42].

3. Incorporation of Novel Agents

It is impossible to consider the future of B-ALL therapy without considering how new agents may be integrated. Recent attempts to intensify therapy with either higher doses of conventional chemotherapy [3,7] or the addition of cytotoxic agents, such as etoposide or clofarabine, have been unsuccessful [43,44]. This suggests that further improvements to therapy for patients with high-risk disease will require novel strategies. Current trials for patients with B-ALL are exploring integration of the CD22-targeted antibody–drug conjugate inotuzumab as part of backbone therapy for higher-risk patients (NCT04307576 and NCT03959085). Other studies are evaluating the CD3-CD19 bispecific antibody construct blinatumomab to attempt to improve therapeutic outcomes for both NCI standard-risk (NCT03914625) and high-risk populations (NCT03117751 and NCT03643276). While current studies are using these agents to either supplement existing therapy or replace weeks of low-intensity maintenance therapy with blocks of these novel agents, future studies are likely to replace intensive blocks of conventional therapy with these agents in an attempt to reduce the long-term toxicities observed in survivors. Further experience with newly approved formulations of asparaginase, including Calaspargase pegol and asparaginase erwinia chrysanthemi (recombinant), will also reveal whether toxicity profiles with these agents differ from historical formulations of this pediatric backbone agent.

For patients with relapsed B-ALL, in recent years, the activity of the novel agents described above has been demonstrated. These agents have shown high response rates even in children with refractory disease after relapse. Data from both clinical trials and compassionate access use of inotuzumab have consistently demonstrated its activity as a single agent in both high-bulk and lower disease burden states [45,46,47]. Blinatumomab has also been incorporated into multiple treatment backbones, with moderate efficacy in bulk disease and excellent activity in patients in morphologic remission with low disease burden [9,48,49,50]. Recent studies on BCL-2 homology domain 3 (BH3) inhibitors have provided hope for this patient population. Data obtained from adults and children have suggested that the BCL-2 inhibitor venetoclax [51] and the BCL2-2, BCL-xL, and BCL-w inhibitor navitoclax [52] are active both preclinically and in patients with relapsed ALL. For patients with KMT2A-rearranged ALL, there is hope that menin inhibitors may improve outcomes for this high-risk population [53,54]. Given the high failure rates of conventional reinduction strategies such as the UKALL R3 regimen, [8,55] we anticipate that current and future studies will evaluate optimal ways to integrate novel agents in both the reinduction and consolidation of children with relapsed B-ALL. New strategies will also be needed for patients with extramedullary relapse, as they appear not to respond well to either blinatumomab or inotuzumab.

4. Cellular Therapies

Cellular therapy is often used to treat pediatric patients with high-risk B-ALL who have either failed traditional chemotherapy regimens or who are deemed to be at a high risk of subsequent disease relapse despite an initial response to therapy [56,57,58]. Given the presumed chemo-resistance of these patient subsets, allogeneic HCT offers an alternative treatment approach by which the use of donor cells provides an immunotherapeutic graft-versus-leukemia (GVL) effect. Outcomes after allogeneic HCT have improved over time, owing to advances in both transplantation and supportive care approaches, which, in turn, contribute to improved leukemia-free survival (LFS) as well as a reduction in transplant-related mortality (TRM) [59,60,61,62]. This includes progress related to conditioning regimens, donor typing, graft manipulation, expansion of access to alternative donors, graft-versus-host disease (GVHD) prophylaxis and treatment, and infection prevention and control [62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78]. Furthermore, advances in leukemic diagnostics, genetic subtyping, and leukemia-directed therapies have decreased the number of patients proceeding to HCT and, for those who are deemed candidates, have enhanced the depth of remission which patients achieve prior to HCT [9,48,49]. Recent findings that HTS-detectable MRD is associated with inferior outcomes in patients who are FC MRD-negative after tisagenlecleucel highlights the impacts of these changes and the role of transplant in these settings [32]. Importantly, with the continued advent of new treatment strategies in both the up-front and relapse settings, there are likely patients that can achieve long-term remission without the use of HCT. Given the risk of morbidity and mortality with current HCT approaches, it will be particularly prudent to continue to re-evaluate indications for transplant in this patient population.

Despite these advances in HCT, there is much work ahead to improve LFS for patients with refractory B-ALL pre-HCT, as outcomes remain dismal in this cohort, primarily due to high rates of disease relapse after HCT [79,80]. In addition to efforts to reduce the disease burden pre-HCT through further discovery and incorporation of novel agents in relapse therapy regimens, it is imperative that research continues to focus on immunotherapeutic strategies to promote GVL while not increasing the risk of GVHD. This includes exploration into graft manipulation and donor sources, the use of selective cell add-back (e.g., NK cells, Tregs), post-HCT pre-emptive therapies (e.g., blinatumomab, ABL tyrosine kinase inhibitors, and/or cytokine support), and GVHD prophylaxis approaches. For this high-risk patient population, combinatorial therapies will likely be needed to overcome leukemia resistance mechanisms while minimizing overlapping toxicity profiles.

Beyond HCT, CD19-redirected CAR T-cell therapy has transformed treatment options for pediatric patients with historically incurable relapsed/refractory B-ALL, inducing remission in the majority of patients treated. These outcomes have been consistent across numerous trials despite differences in CAR construct, clinical trial design, patient inclusion/exclusion criteria, lymphodepleting regimens, and decisions about post-CAR T-cell therapy, such as consolidative allogeneic HCT [81,82,83,84,85,86]. Importantly, real-world outcomes using the FDA-approved product tisagenlecleucel have continued to mirror those seen in the initial clinical trials [83,87,88,89,90].

Despite these successes, relapse after treatment with CD19-CAR T-cell therapy continues to be a significant clinical challenge, with approximately 50% of treated patients experiencing subsequent disease relapse across different products. This includes a subset of patients with antigen escape, during which the CD19 target is lost [83,91,92]. Importantly, outcomes for those patients with initial non-response or subsequent relapse are quite poor [93]. While close monitoring for the loss of B-cell aplasia and improvements in post-CAR disease monitoring, such as next-generating sequencing MRD measurement, provides some further information to guide treatment decisions after CD19-CAR T-cell therapy, relapse still occurs [83,88,89,90]. For some patients, CAR T-cell therapy is, therefore, used as a bridge to a planned consolidative allogeneic HCT [89,94,95]. However, given the morbidity and mortality risk of HCT, there is a critical need to continue to improve CAR T-cell approaches to promote sustained remission, as well as efforts to better define high-risk patients prior to CD19-CAR T-cell treatment. We anticipate that such work will accelerate, as recent work has already begun to identify patients with lower response rates to existing CAR T-cell therapies [96].

A primary area of investigation to improve outcomes after CAR T-cell therapy for pediatric patients with B-ALL includes the evaluation of alternative antigens, primarily CD22 alone or in combination with CD19 [85,89]. Several trials are currently underway in the pediatric space, with promising interim results. Additionally, efforts to promote persistence and limit immune rejection of infused autologous CAR T-cell products include reinfusion [85,89,97,98,99,100], humanized CAR T cell products [101], the addition of PD-1 inhibitors [102], gene editing of autologous T cells, and allogeneic approaches [103,104,105]. Data from prior trials also adds evidence which can be used to explore T-cell biology further for potential modifiable factors that may augment treatment responses, such as reducing DNMT3A-driven methylation resulting in T-cell exhaustion [106] or new CAR constructs which utilize priming antigens to control CAR expression [107]. An ongoing commitment from laboratory scientists to identify strategies to improve CAR T-cell therapy, from clinician scientists to evaluate these approaches in early phase trials, and from industry and regulatory partners to allow broad access to such new treatments will be needed.

Furthermore, traditional risk factors for treatment failure, both in newly diagnosed and relapsed pediatric patients, are not consistently replicated after treatment with CD19-CAR T-cell therapy, suggesting that this treatment may overcome the excess risk observed in some populations. This includes subgroups with high-risk cytogenetics [108], Down syndrome [109], infants [110,111], and extramedullary disease [112]. Encouraged by the ability of CD19-CAR T cells to eradicate leukemic cells in the cerebrospinal fluid of patients with relapsed B-ALL, several studies tested this approach in the treatment of isolated extramedullary relapse [112,113,114,115]. The preliminary results are encouraging, especially for patients with testicular relapse. CAR T-cell therapy could become a therapeutic option for patients with extramedullary relapse, sparing them from receiving local irradiation. This highlights the potential of CAR T-cell therapies to expand treatment options and circumvent previously defined high-risk characteristics, thereby re-defining our approaches for these patients. Crucially, while there is a growing body of literature reporting on such potential factors, continued evaluation of such therapies across larger patient cohorts is necessary to further elucidate the associations, including those patient groups treated with commercially available and novel experimental products. This will allow for more prudent risk stratification of pediatric patients with relapsed/refractory B-ALL, informing choices of CAR T-cell therapy, post-infusion monitoring, and/or treatment strategies.

5. Conclusions

While serial clinical trials over the last 60 years have transformed B-ALL from an almost universally fatal disease to one in which more than 9/10 children can expect to be long-term survivors, much effort is still needed to improve treatment for this common pediatric cancer. Recent progress in more sensitive disease detection methods and the availability of novel cellular and chemotherapy options provides hope that further improvements to the cure can be achieved, while simultaneously decreasing the short- and long-term therapy toxicity levels. We anticipate that trials over the next decade will continue to transform the care in this field, with benefits to thousands of children and their families annually.