Key points

Introduction

The remaining challenges for childhood cancer research are best understood in the context of past advances. Treatment of childhood cancer was one of the important success stories of twentieth century medicine as exemplified by the conversion of pediatric acute lymphoblastic leukemia (ALL) from an incurable disease in the 1950s to one in which more than 90% of children survived 5 years from diagnosis, with most of these children cured of their leukemia. Other cancers also have 5-year survival rates approaching or exceeding 90%, including Wilms tumor, non-Hodgkin lymphoma (NHL), Hodgkin lymphoma, and germ cell tumors. Importantly, the decline in childhood cancer mortality that began in the 1960s continued through the first decade of the twenty-first century. Research advances averted more than 45,000 childhood cancer deaths from 1975 to 2010.

Despite the successes in identifying effective treatments for many children with cancer, approximately 2000 children and adolescents die of their disease each year in the United States. Fig. 1 shows the distribution of childhood cancer mortality for children and adolescents, highlighting the contribution of leukemias, brain cancers, and neuroblastoma in younger children and the contribution of leukemias and brain cancers, along with sarcomas and lymphomas, in adolescents. Additionally, for some cancers, progress has been very limited (eg, diffuse intrinsic brainstem gliomas, high-grade gliomas, and metastatic sarcomas). Beyond the number of children who die each year, there is also the burden of long-term morbidity that diminishes quality of life for some childhood cancer survivors.

Patterns of mortality for children and adolescents younger than 15 years and 15 to 19 years for 2007 to 2010. CNS, central nervous system; Oth Leuk, other leukemia.

( Adapted from Smith MA, Altekruse SF, Adamson PC, et al. Declining childhood and adolescent cancer mortality. Cancer 2014;120(16):2500; with permission.)

The challenge for the future is to discover and implement new strategies that will allow successful treatment of those for whom current therapeutic approaches are suboptimal, either because of insufficient efficacy or because of the damage that the treatments cause to critical normal tissues, resulting in acute morbidity and long-term disability. In addressing this challenge, it is important to acknowledge that most of the anticancer drug and biologic products that will be studied in the context of clinical trials in children will be ones initially developed for adult cancers. Even so, it is critical that pediatric oncologists prioritize these agents independently of their utility for adult cancers, as both the biology and the goals of treatment generally differ between childhood and adult cancers. For example, the primary goal in treating childhood cancers is cure, not palliation, whereas for many adult cancers sustained stable disease aimed at palliation is an important objective. Agents and treatment regimens that only slow tumor growth and prevent disease progression for a finite period may be valuable as adult cancer treatments, assuming that they prolong survival while allowing acceptable quality of life ; however, for children, temporarily delaying disease progression is at best a modest success. This critical distinction between the relative benefit of cure versus palliation for children and adults with cancer, coupled with the differences in etiology and biology of pediatric and adult cancers, has implications both in terms of the cellular pathways targeted for intervention and in terms of clinical trial design. Moreover, it highlights the need for pediatric cancer prioritization decisions to focus on the biology of the cancers and on the specific needs of the children afflicted with these cancers. We discuss factors that will play key roles in guiding pediatric oncologists as they select lines of research to pursue in their quest for more effective treatments for children with cancer.

Introduction

The remaining challenges for childhood cancer research are best understood in the context of past advances. Treatment of childhood cancer was one of the important success stories of twentieth century medicine as exemplified by the conversion of pediatric acute lymphoblastic leukemia (ALL) from an incurable disease in the 1950s to one in which more than 90% of children survived 5 years from diagnosis, with most of these children cured of their leukemia. Other cancers also have 5-year survival rates approaching or exceeding 90%, including Wilms tumor, non-Hodgkin lymphoma (NHL), Hodgkin lymphoma, and germ cell tumors. Importantly, the decline in childhood cancer mortality that began in the 1960s continued through the first decade of the twenty-first century. Research advances averted more than 45,000 childhood cancer deaths from 1975 to 2010.

Despite the successes in identifying effective treatments for many children with cancer, approximately 2000 children and adolescents die of their disease each year in the United States. Fig. 1 shows the distribution of childhood cancer mortality for children and adolescents, highlighting the contribution of leukemias, brain cancers, and neuroblastoma in younger children and the contribution of leukemias and brain cancers, along with sarcomas and lymphomas, in adolescents. Additionally, for some cancers, progress has been very limited (eg, diffuse intrinsic brainstem gliomas, high-grade gliomas, and metastatic sarcomas). Beyond the number of children who die each year, there is also the burden of long-term morbidity that diminishes quality of life for some childhood cancer survivors.

Patterns of mortality for children and adolescents younger than 15 years and 15 to 19 years for 2007 to 2010. CNS, central nervous system; Oth Leuk, other leukemia.

( Adapted from Smith MA, Altekruse SF, Adamson PC, et al. Declining childhood and adolescent cancer mortality. Cancer 2014;120(16):2500; with permission.)

The challenge for the future is to discover and implement new strategies that will allow successful treatment of those for whom current therapeutic approaches are suboptimal, either because of insufficient efficacy or because of the damage that the treatments cause to critical normal tissues, resulting in acute morbidity and long-term disability. In addressing this challenge, it is important to acknowledge that most of the anticancer drug and biologic products that will be studied in the context of clinical trials in children will be ones initially developed for adult cancers. Even so, it is critical that pediatric oncologists prioritize these agents independently of their utility for adult cancers, as both the biology and the goals of treatment generally differ between childhood and adult cancers. For example, the primary goal in treating childhood cancers is cure, not palliation, whereas for many adult cancers sustained stable disease aimed at palliation is an important objective. Agents and treatment regimens that only slow tumor growth and prevent disease progression for a finite period may be valuable as adult cancer treatments, assuming that they prolong survival while allowing acceptable quality of life ; however, for children, temporarily delaying disease progression is at best a modest success. This critical distinction between the relative benefit of cure versus palliation for children and adults with cancer, coupled with the differences in etiology and biology of pediatric and adult cancers, has implications both in terms of the cellular pathways targeted for intervention and in terms of clinical trial design. Moreover, it highlights the need for pediatric cancer prioritization decisions to focus on the biology of the cancers and on the specific needs of the children afflicted with these cancers. We discuss factors that will play key roles in guiding pediatric oncologists as they select lines of research to pursue in their quest for more effective treatments for children with cancer.

Genomic alterations as therapeutic guideposts

One essential line of research for identifying more effective treatment strategies is understanding in detail the genomic alterations that provide the blueprint for the growth and survival signaling pathways of childhood cancers. These genomic alterations highlight the genes that the cancers are most dependent on, whether for their oncogenic driver effect or for their tumor suppressor role. Oncogenes with genomic alterations have proven among the most useful guideposts for identifying therapeutic targets, as illustrated by the success of imatinib for BCR-ABL leukemias (chronic myeloid leukemia and Philadelphia-positive [Ph+] ALL) and the success of crizotinib for ALK -rearranged non–small cell lung cancer. The success of imatinib when added to standard chemotherapy for children with Ph+ ALL is particularly informative. Single-agent imatinib induces remissions of relatively short duration in Ph+ ALL, whereas standard chemotherapy is effective for a minority of children (approximately 30%). However, the combination of imatinib and standard chemotherapy was able to induce and maintain long-term remission in approximately 70% of children with Ph+ ALL.

A decade ago it was possible to hope that targetable oncogenes might be identified in a high percentage of childhood cancers and that the imatinib paradigm described previously could be broadly applied to other childhood cancers. At this point, thousands of childhood cancer specimens have been sequenced so that the vast majority of recurring mutations have now been identified. Targetable activated oncogenes have been identified, including the NPM-ALK fusion gene for anaplastic large cell lymphoma, ALK point mutations for a subset of neuroblastoma, BRAF genomic alterations for pediatric gliomas, Hedgehog pathway mutations for a subset of medulloblastoma, and ABL family genes activated by translocation in a subset of Ph-like ALL. However, these examples represent a small minority of all childhood cancers, making it clear that most childhood cancers do not have recurring mutations in genes that are, at the present time, considered targetable.

Two approaches to therapeutically targeting the “untargetable” warrant mention. One is the concept of identifying targetable susceptibilities created by untargetable genomic alterations. This concept is illustrated by the activity of EZH2 inhibitors in rhabdoid tumors. SMARCB1 loss of function through deletion or mutation is the sole recurring genomic alteration in rhabdoid tumors. EZH2 is a member of the Polycomb Repressor Complex 2 (PRC2) that mediates gene silencing through catalyzing trimethylation of histone 3 lysine 27 (H3K27) at the promoters of target genes. Mice with conditional loss of SMARCB1 in their T cells rapidly develop T-cell lymphomas, but tumor development is completely suppressed by concomitant loss of EZH2. Small molecule inhibitors of EZH2 have been developed and have entered clinical evaluation. Treatment of rhabdoid tumor xenografts with an EZH2 inhibitor led to dose-dependent tumor regression, providing evidence for the potential clinical utility of EZH2 inhibitors for cancers with SMARCB1 loss of function. Another example of targetable susceptibilities created by untargetable genomic alterations is the requirement of mixed lineage leukemia (MLL)-rearranged leukemias for the DOT1L methyltransferase. A small molecule inhibitor of DOT1L induced complete regressions in a xenograft model of MLL leukemia, and this agent has entered clinical evaluation.

A second approach to targeting the untargetable is applying medicinal chemistry and high-throughput screening methods to identify small molecule inhibitors of pediatric oncogenes. This strategy is illustrated by efforts at developing small molecule inhibitors of EWS-FLI1 activity that resulted in development of YK-4-279, a small molecule that blocks EWS-FLI1 from interacting with RNA Helicase A. Other pediatric oncogenes that are candidates for targeting include the PAX-FKHR fusion proteins of alveolar rhabdomyosarcoma and MYCN, which is amplified in high-risk neuroblastoma in children older than 18 months.

Although genomic alterations are reliable therapeutic guideposts for many targeted agents, the role of the tissue of origin should not be overlooked as a potential guidepost for some targeted agents. As an example, proteasome inhibitors are effective for patients with multiple myeloma, even though mutations in proteasome subunits are exceedingly rare. The susceptibility of myeloma to proteasome inhibition likely relates to the high level of synthesis of immunoglobulins in myeloma cells, which leads to a dependence on proteasome function to process the resulting elevated levels of unfolded proteins. Similarly, inhibitors of PI3K delta are highly active in malignancies of mature B cells, not because of PIK3CD mutations, but because of the dependence of these mature B cells on signaling through the B-cell receptor. The favorable therapeutic impact of engaging the glucocorticoid receptor in ALL cells is a pediatric example of the importance of tissue-of-origin effects.

Molecularly defined disease subtypes

A consequence of the detailed molecular characterization of childhood cancers is the recognition that single disease entities actually represent multiple clinically and biologically distinctive subtypes. For example, analysis of gene expression profiles of B-ALL cases identified 8 distinctive subtypes, 1 of which had similar characteristics as Ph+ ALL but lacked the BCR-ABL fusion gene, Further investigation of this “Ph-like” ALL subset showed that these cases in turn possess a range of genomic alterations, with most having an alteration in genes involved in growth factor signaling. These alterations include potentially therapeutically relevant fusion genes involving tyrosine kinases (eg, PDGFRB , ABL1 , and CSF1R ), as well as well as genomic alterations involving CRLF2 , JAK family members, and RAS pathway alterations.

Detailed investigation of medulloblastoma cases has identified 4 molecular subtypes, each with a distinctive constellation of genomic alterations as well as distinctive demographic and prognostic characteristics. One subtype, the sonic hedgehog (SHH) group, is characterized by mutations in the SHH pathway. The SHH pathway can be activated by genomic alterations in a number of genes, including PTCH1 , SUFU , GLI2 , MYCN , and SMO . Only cases with “upstream” mutations in the SHH pathway (eg, PTCH1 and SMO ) are susceptible to inhibition by currently available SHH pathway inhibitors that block SMO action (eg, vismodegib and sonidegib). Patients within the SHH subtype of medulloblastoma show distinctive genomic profiles by age, with infants having primarily either PTCH1 or SUFU mutations, older children having PTCH1 mutations or GLI2 amplification, and adults having primarily PTCH1 mutations. For the pediatric age range, up to 50% of cases have lesions downstream of SMO that are inherently nonresponsive to these agents. This example highlights the complexities of targeted therapy development in children, even when a targetable oncogenic pathway is activated in a specific patient population.

Some have proposed that the molecular characterization of cancer heralds the end of the era of histology-defined treatment and the move to an era in which specific genomic alterations rather than histology will define treatment. For childhood cancers, there is reason for a more conservative approach to research strategy in which molecular characterization complements, but does not replace, histologic classification of cancers. One reason for this conservative approach is the remarkable relationship between specific genomic alterations and specific cancer types, as illustrated by the finding of H3F3A K27M mutations in midline high-grade gliomas of children but in virtually no other cancers. Similarly, BRAF mutations do not occur randomly across childhood cancers, but are found primarily among cases of low-grade gliomas. A second reason for skepticism is that the therapeutic implications of genomic alterations can be cell context dependent, as illustrated by the high activity of BRAF inhibitors for patients with melanoma and BRAF V600E mutations, but their low activity in patients with colorectal cancer with the same mutation. A final note of caution that is particularly relevant for childhood cancers is that the development pathway for targeted agents that show single-agent activity will likely be (at least initially) through their integration with standard therapy, as illustrated by the imatinib example for Ph+ ALL described previously. To the extent that different histologies have different standard treatments, the development of targeted agents will be accordingly segregated by histology.