The Zero Childhood Cancer Program’s multi-platform sequencing approach identified molecular alterations in 94% of a cohort of 247 pediatric patients with high-risk cancers, which has enabled more-precise diagnoses and alternative therapeutic recommendations.
Cancer is a leading cause of death and treatment-related morbidity in children and young adults worldwide1,2. While there have been drastic improvements in survivorship for acute lymphoblastic leukemia, similar progress for other childhood cancers has been modest. Indeed, several childhood solid tumors remain intractable, while metastatic disease is hard to treat and is associated with a poor outcome. Next-generation sequencing (NGS) technologies may help improve the detection of specific molecular alterations in rare childhood tumors for precise therapeutic targeting. In this issue of Nature Medicine, Wong et al. present one of the first initiatives integrating multiple NGS platforms to inform targeted therapies for pediatric cancer, the Zero Childhood Cancer Program (ZERO), Australia’s first national precision-medicine program for pediatric cancer, and provide initial promising results3.
The limited success in the treatment of childhood and young adult solid cancers can be attributed to the lack of large cohorts, limited longitudinal studies across the course of the disease, the absence of reliable biomarkers and the few available targeted therapies specific for children with cancer. Moreover, pediatric tumors are mainly mesenchymal in origin, while they are mostly epithelial in adults, and recent NGS studies have uncovered a previously unappreciated molecular diversity in several pediatric cancers, as exemplified in central nervous system (CNS) tumors4. In addition, pediatric cancers display distinct genetic alterations and a lower mutational burden, indicative of specific biological mechanisms and novel oncogenic pathways driven by rare events in these neoplasms5,6,7. These observations have prompted the establishment of large-scale precision-medicine programs worldwide to improve outcomes in childhood cancer.
Here, Wong et al. carry out one such nationwide study3. They present integrated data from a comprehensive genomic, transcriptomic and epigenomic analysis of 252 tumor samples from children, adolescents and young adults under 21 years of age with refractory or hard-to-treat tumors (<30% 5-year survival expected) recruited from all major childhood cancer centers in Australia (Fig. 1). Each tumor and the associated patient’s germline DNA were profiled by whole-genome sequencing (WGS), the tumor transcriptome was investigated by RNA sequencing (RNA-seq), and CNS tumors were further analyzed through the use of DNA-methylation arrays. This multi-platform sequencing approach enabled the identification of an extensive number of molecular alterations in tumors, including single-nucleotide variants, structural variants, copy-number variants and gene fusions. Notably, the authors identify a total of 1,023 reportable tumor-driving alterations, with nearly 94% of patients included in this study having at least one reportable alteration. On the basis of actionable molecular alterations, a total of 134 patients (67%) received therapeutic recommendations, half of them relying on tier 1 or tier 2 levels of evidence (alterations with supporting clinical evidence in the same cancer or different cancer, respectively). In addition, NGS helped rectify several diagnoses, especially in CNS tumors, as previously shown4,8,9.
The use of orthogonal NGS approaches is a strength of the ZERO initiative. Their results show that combining results from distinct platforms increases confidence in the variants identified, such as by validating the impact of splice-site variants in RNA-seq or confirming gene-expression outliers by analysis of copy-number variants. Moreover, joint detection by RNA-seq and WGS of aberrant fusion transcripts increases the likelihood that these are pathogenic. Approximately 40% of the reportable variants the authors identify are supported by both WGS and RNA-seq, while a notable 60% of the variants are identified through a single technology. The expression data the authors acquire by RNA-seq to identify potential therapeutic targets is beneficial, as ~63% of aberrantly expressed genes could not be explained by the coding genome. For example, high expression of CTLA4 (which encodes the immunomodulatory receptor CTLA-4) is identified only in RNA-seq and suggests the presence of an immune infiltrate in a tumor and thus that this tumor might respond to immune-checkpoint inhibition.
The authors’ approach further illustrates the power of WGS over targeted sequencing in identifying pathogenic and druggable mutations, despite the associated cost, complex and lengthy data analyses, and data-storage challenges. WGS helps them to identify complex chromosomal rearrangements, small and large copy-number variants, and variants in regulatory elements, which are traditionally harder to detect by other methods. For example, they identify several novel gene fusions, including a BRD4–LEUTX fusion that could potentially be targeted through the use of epigenetic inhibitors. Furthermore, using novel analysis methods, they shed light on complex events that lead to fragmented chromosomes, such as one leading to a TP53–SUZ12 fusion, which provides unique insights into tumor biology.
Notably, the authors’ WGS proves instrumental for identifying complex germline events. Wong et al. show that the rate of pathogenic germline variants in their cohort of patients with high-risk tumors is 16% (ref. 3), a frequency much higher than the previously suspected ~8.5% (ref. 10). Only 35% of families were aware of a germline risk variant before the comprehensive profiling performed by the ZERO initiative. While this may represent a bias of this specific population, these results call for further investigations to assess the real baseline predisposition in childhood cancers.
Precision-oncology initiatives in hard-to-treat pediatric cancers are increasing in developed countries11. In the USA, several large centers are carrying out such studies, following the TARGET initiative of the Children’s Oncology Group and the US National Cancer Institute (Therapeutically Applicable Research to Generate Effective Treatments; http://ocg.cancer.gov/programs/target), open in select pediatric centers able to perform phase 1/2 clinical trials. The TARGET initiative aims to inform researchers of clinically actionable targets in sequenced tumors and promote enrollment in pediatric basket trials. In Canada, the PROFYLE initiative (Precision Oncology for Children and Young People) uses various NGS platforms and clinical trial approaches similar to those of the TARGET trial. In Europe, initiatives such as INFORM (Individualized Therapy for Relapsed Malignancies in Childhood) carry out DNA-methylation and various NGS analyses to match patients with available clinical trials. As more than 70% of pediatric patients are usually enrolled in clinical trials, compared with less than 10% of adult patients, these programs will rapidly achieve recruitment, especially given the patient community’s eagerness to participate in NGS-based initiatives. These growing programs are essential for assembly of the larger patient cohorts required for development of the invaluable clinical and NGS datasets needed to improve knowledge of pediatric oncology. Providing open access to data and timely results to the scientific community will help systematic and thorough cataloguing of relevant molecular alterations, including variants of unknown significance, and will better identify targeted therapies on the basis of clinical outcome. These studies will prove an invaluable resource for identifying molecular-based tumor subtypes and biomarkers predictive of drug response and will thus enable the development of a broader spectrum of therapeutic options, while guiding the use of much-needed combination therapies.
In conclusion, Wong et al. demonstrate the feasibility and benefits of integrating genomic and epigenomic data in real time to direct treatment decisions for pediatric patients with high-risk cancers through a prospective national precision-medicine program3. As initiatives such as ZERO grow worldwide, the choice among the ever-evolving NGS technologies should be made carefully. It should be based on expertise, tumor type, and the state (fresh, frozen or fixed) and amount (tiny biopsy or full resection) of material available for analysis, accepting that no individual NGS platform will provide all answers and some NGS approaches are complementary. Many additional steps are needed to ensure success in these initiatives. They require coordination among people involved in patient care to streamline the acquisition of informed consent, the gathering of material in close collaboration with surgeons, pathologists, oncologists and research teams, timely sequencing and analysis of NGS results, accurate reporting, enrollment in clinical trials on the basis of target identification, and the collection of results from these trials to feed databases for future reference. As these technologies transition to become the standard of care, the perceived high cost of NGS needs to be weighed with the expected clinical benefit, a premise supported by this study3 and similar reports demonstrating faster and more-reliable diagnoses, improved prognosis, familial cancer risk assessment and better access to meaningful targeted therapies.
Steliarova-Foucher, E. et al. Lancet Oncol. 18, 719–731 (2017).
Cunningham, R. M., Walton, M. A. & Carter, P. M. N. Engl. J. Med. 379, 2468–2475 (2018).
Wong, M. Nat. Med. https://doi.org/10.1038/s41591-020-1072-4 (2020).
Capper, D. et al. Nature 555, 469–474 (2018).
Vogelstein, B. et al. Science 339, 1546–1558 (2013).
Gröbner, S. N. et al. Nature 555, 321–327 (2018).
Schwartzentruber, J. et al. Nature 482, 226–231 (2012).
Sturm, D. et al. Cell 164, 1060–1072 (2016).
Sturm, D. et al. Cancer Cell 22, 425–437 (2012).
Zhang, J. et al. N. Engl. J. Med. 373, 2336–2346 (2015).
DuBois, S. G., Corson, L. B., Stegmaier, K. & Janeway, K. A. Science 363, 1175–1181 (2019).
The authors declare no competing interests.
About this article
Cite this article
Hadjadj, D., Deshmukh, S. & Jabado, N. Entering the era of precision medicine in pediatric oncology. Nat Med 26, 1684–1685 (2020). https://doi.org/10.1038/s41591-020-1119-6