← Back to Home

Identifying High-Risk Cells in Tumors: Neuroblastoma Study Finds a New Gene Signature

Cancer cells that are difficult to distinguish from normal tissue may hold clues to differences in children’s treatment outcomes. Researchers developed an identification signature and experimental models from patient tumors, providing a more reliable starting point for investigating treatment resistance.

By SURL BioNews

Children with the same diagnosis of neuroblastoma can have vastly different treatment outcomes. To understand this gap, researchers first need a clear picture of the cancer cells within tumors. St. Jude Children’s Research Hospital and its collaborators have now identified a gene expression signature that can distinguish mesenchymal malignant cells that were previously difficult to identify; patients with stronger expression of this signature also had poorer survival outcomes. The study was published in *Cancer Cell* on October 8.

The difficulty in identifying these cells stems from similarities between cancer cells and normal cells. Neuroblastoma cells can exist in two states, adrenergic and mesenchymal, with the latter thought to be associated with treatment resistance. However, tumors also contain normal stromal cells, which can complicate identification. According to the paper’s abstract listed by the Broad Institute, existing gene signatures derived from laboratory cell lines cannot effectively identify mesenchymal cancer cells in patient tumors.

The team analyzed 54 tumors from 50 patients, combining single-cell and single-nucleus RNA sequencing, bulk RNA sequencing, and spatial transcriptomic and proteomic analyses that retain information about tissue location. These methods allowed researchers to read cells’ molecular signals while also seeing their distribution within tumors, clarifying the distinct regions occupied by cancer cells and nonmalignant immune and stromal cells.

The new identification signature was derived from early-passage patient-derived xenograft models grown at the corresponding anatomical sites in mice. According to St. Jude, normal stromal cells from patient tumors do not continue growing in these models, helping researchers isolate the signals from malignant cells. The researchers then cross-checked their findings using organoids, cell lines, electron microscopy, and chromatin analyses, among other methods, confirming differences between the two cell states in molecular features, morphology, and function.

The signature was also validated using independent RNA sequencing data from patient tumors. The paper’s abstract shows that higher mesenchymal gene expression was associated with poorer event-free and overall survival. St. Jude’s press release and the content republished by Medical Xpress state that the new signature significantly improved outcome prediction, while the existing cell-line signature did not show the same effect. However, this remains evidence of an association and does not establish that these cells directly cause poorer treatment outcomes.

Another result of the study is a set of patient-derived models containing different proportions of mesenchymal and adrenergic cells, which can be requested through the Childhood Solid Tumor Network. The European Genome-phenome Archive also contains 219 RNA sequencing samples associated with this study, with data access subject to approval by an access committee. This figure represents the number of sequencing samples and cannot be interpreted as 219 patients.

These tools give researchers an opportunity to test mechanisms of treatment resistance and potential therapeutic vulnerabilities more accurately, but key questions remain before clinical use. The available abstract and press releases do not provide the magnitude of the predictive improvement, thresholds for use, or prospective validation results needed to assess clinical benefit. Nor has it been demonstrated that adjusting treatment based on this signature improves survival. The immediate advance is making cancer cells that were difficult to identify into research subjects that can be tracked and studied experimentally.

References

  1. St. Jude Children's Research Hospital
  2. Broad Institute
  3. European Genome-phenome Archive
  4. Medical Xpress