By Dibash Kumar Das, PhD
Pediatric solid tumors make up around 40 percent of all childhood cancers and they can occur in many areas of the body. Genomic profiling has been used to aid in diagnosis and determine the prognosis for pediatric cancers and guide treatment in some cases. However, pediatric tumors have different genetic characteristics compared to adult tumors, which means that specific tests are needed to diagnose and treat them.
While adult tumors are often caused by mutations, pediatric tumors are more frequently characterized by copy number alterations (CNAs), gene fusions, and other large-scale genetic rearrangements. Recent studies have shown that using cell-free DNA (cfDNA) from plasma can help detect genetic changes in pediatric tumors (Cancer Discov 2022; https://doi.org/10.1158/2159-8290.CD-21-1136), but more research is needed to develop clinical tests to diagnose and monitor a variety of pediatric solid tumors.
Now, scientists at Children's Hospital Los Angeles (CHLA) have created a liquid biopsy (LB) for pediatric solid tumors that can help diagnose specific types of cancer when traditional biopsies are not feasible (npj Precis Onc 2023: https://doi.org/10.1038/s41698-023-00357-0). The study will investigate whether plasma-based LB and low-pass whole genome sequencing (LP-WGS) might support the clinical diagnosis and prognosis of pediatric solid tumors, monitor therapy response, and detect early evidence for relapse. The researchers combined LP-WGS with targeted sequencing of cfDNA from plasma to identify CNAs, mutations, and gene fusions linked with pediatric solid tumors.
A significant aspect of this study was that the amount of sample required for testing was much smaller than that required for LB studies in adults, since infants and young children have smaller blood volumes, and the assays were scaled down to adjust for this. With LP-WGS, many samples can be loaded onto a sequencer at once, allowing for low-cost testing. Moreover, the assay could be used across pediatric tumors and use different sources for LB samples, such as aqueous humor, cerebral spinal fluid, and blood samples.
A total of 143 plasma samples were analyzed, including 73 from patients with newly diagnosed or recurrent malignant bone or soft tissue sarcomas, as well as germ cell, hepatic, thyroid, or renal tumors, and 19 from non-cancer control patients. Researchers collected cfDNA at diagnosis, during and after therapy, and/or at relapse, which was isolated from plasma samples.
Specifically, the assay detected that 70 percent (26 out of 37) of patients had circulating tumor DNA (ctDNA) in the diagnostic samples and 43 percent (10 of 19) in relapse samples. The LP-WGS test also detected CNAs in 67 percent (18 of 27) of patients with localized disease and 80 percent (8 of 10) of patients with metastatic disease. This data shows that, even with previous therapy, patients could still have enough ctDNA in their plasma to be found by LP-WGS. The control group did not have detectable somatic CNAs. Notably, this study is one of the few to demonstrate the applicability of ctDNA detection in pediatric solid tumor patients with localized disease.
Next, the researchers investigated whether the test could identify CNAs associated with specific types of tumors. In osteosarcoma patients, the assay identified genomic instability, a hallmark of the disease, in 53 percent of patients out of 17 tested. Among 13 Ewing sarcoma patients, nine patients had detectable CNAs with seven patients showing a gain of chromosome 8, an abnormality frequently observed in this cancer type. The assay also detected CNAs in 75 percent of newly diagnosed renal tumor patients and in all five relapsed renal tumor patients.
Next, mutations identified in tumor samples using a next-generation sequencing (NGS) panel, OncoKids, were also observed in 14 out of 26 plasma samples through the analysis of ctDNA using LP-WGS. A total of 16 hotspot mutations in 14 patients were detected, including alterations in PIK3CA, TP53, and CTNNB1, demonstrating the sensitivity of LP-WGS, the low but significant number of mutations present in pediatric solid tumors. Lastly, the researchers developed a hybridization-based capture panel to detect EWSR1 and FOXO1 fusions, which were detected in 10 out of 12 patients with Ewing sarcoma and 2 out of 2 patients with alveolar rhabdomyosarcoma.
The researchers suggest that a larger cohort size is needed to confirm the applicability of fragment size analysis in detecting ctDNA and distinguishing tumor types in pediatric cancers. They also propose that validating the use of LB assays in larger cohorts of patients with pediatric solid tumors, combined with clinical biomarkers such as alpha-fetoprotein, could lead to a non-invasive means of molecular diagnosis and monitoring patients from diagnosis through treatment and recurrence. Serial studies using NGS-based LB assays could help in understanding how they can be employed for this purpose.
For more insights into the study, Oncology Times connected with Jaclyn Biegel, PhD, FACMG, senior author and Division Chief of Genomic Medicine and Director of the Center for Personalized Medicine at Children's Hospital Los Angeles, as well as Professor of Clinical Pathology in the Keck School of Medicine at the University of Southern California.
Oncology Times: Can you elaborate on the potential clinical applicability of your LB platform to evaluate pediatric patients with a variety of solid tumors, and how it differs from other LB assays available on the market?
Biegel: “The majority of LB assays available on the market today target the most frequent DNA alterations observed in tumors in adults, for example activating mutations in oncogenes in colon cancer and lung cancer. The cfDNA yield from 20 mLs of blood in an adult is more than adequate to perform those assays, but it is not feasible in pediatric patients. Pediatric tumors are more often characterized by tumor-associated patterns of CNAs and gene fusions, and very few mutations. The LB assays we have developed require an input of only 1-5 ng of cfDNA isolated from 50 μL of aqueous humor from patients with eye tumors, or 2-4 μL of cerebrospinal fluid (CSF) or blood from patients with brain or solid tumors. The first clinically validated LB assay at Children's Hospital Los Angeles employs low-pass WGS to detect copy number changes in newly diagnosed patients, as well as patients with recurrent or metastatic disease. Subsequent versions of the assay will employ targeted sequencing to detect mutations and gene fusions."
Oncology Times: How do you ensure that you are accurately distinguishing ctDNA from cfDNA, given the challenge of cfDNA being derived predominantly from hematopoietic cells?
Biegel: “The use of low-pass WGS allows us to identify CNAs in ctDNA that are not seen in the germline (normal cells). The challenge arises when the tumor has a normal copy number profile; for example, in Ewing sarcomas with balanced translocations that result in EWSR1 gene fusions. Targeted sequencing to detect specific gene fusions or mutations will result in higher sensitivity and specificity to detect ctDNA in LB specimens. ctDNA fragments are shorter than fragments derived from normal cells and thus we also used in-silico approaches to enrich for ctDNA content in our analyses."
Oncology Times: The study mentions that the incidence of pediatric solid tumors is low, and the number of histologic, genomic, and clinical subtypes is large. How do you account for this heterogeneity when developing a pan-cancer assay for LB?
Biegel: “The utility of WGS as a basis for the assay ensures that we will capture the majority of CNAs across the entire tumor genome in the same way that we currently employ chromosomal microarray analysis of bone marrow aspirates or tumor tissue to aid in the diagnosis of hematologic malignancies, and brain and solid tumors. In contrast, most of the currently available LB-based tests focus on particular regions in the adult cancer genomes that are likely to demonstrate mutations or specific gains or losses of a particular target gene. To complement the copy number aspect of the LB assay for pediatric cancer, we will develop targeted sequencing approaches that can be used for multiple tumor types and expand the feasibility studies described in the paper for clinical validation with a larger cohort of patients."
Oncology Times: How do you see the use of LBs for pediatric patients with solid tumors evolving in the next 5-10 years? What challenges need to be addressed to make this a widespread clinical tool?
Biegel: “We expect to see rapid growth in the development of LB assays for pediatric cancer over the next several years. We have already demonstrated the utility of our assay as an aid in diagnosis when a tumor biopsy is not feasible, for detection of high-risk genetic features that may guide therapy, and for early detection of relapse prior to the presence of tumor visualized by imaging. Larger studies across the spectrum of tumor types are required to demonstrate the use of LB assays for surveillance in patients who are at risk for the development of cancer due to genetic predisposition factors or who may relapse after treatment. This will be challenging because many of the subtypes of pediatric cancer are, fortunately, very rare and specimens may still be limited.
“We know very little about the factors that influence the yield of ctDNA in plasma and CSF and are, therefore, being conservative in using the total amount of ctDNA isolated from plasma or CSF as a specific biomarker. Clinical assays based on tumor-associated methylation profiles of ctDNA, non-coding RNAs, mitochondrial DNA variants, and proteins are all in development and will allow us to use LBs for a larger number of applications than we have available at present. The relatively rare nature of pediatric cancer may be an advantage as we aim to develop patient-specific assays for clinical use."
Dibash Kumar Das is a contributing writer.