The NHS England 100,000 Genomes Project: feasibility and utility of centralised genome sequencing for children with cancer

Abstract

Background: Whole-genome sequencing (WGS) of cancers is becoming an accepted component of oncological care, and NHS England is currently rolling out WGS for all children with cancer. This approach was piloted during the 100,000 genomes (100 K) project. Here we share the experience of the East of England Genomic Medicine Centre (East-GMC), reporting the feasibility and clinical utility of centralised WGS for individual children locally.

Methods: Non-consecutive children with solid tumours were recruited into the pilot 100 K project at our Genomic Medicine Centre. Variant catalogues were returned for local scrutiny and appraisal at dedicated genomic tumour advisory boards with an emphasis on a detailed exploration of potential clinical value.

Results: Thirty-six children, representing one-sixth of the national 100 K cohort, were recruited through our Genomic Medicine Centre. The diagnoses encompassed 23 different solid tumour types and WGS provided clinical utility, beyond standard-of-care assays, by refining (2/36) or changing (4/36) diagnoses, providing prognostic information (8/36), defining pathogenic germline mutations (1/36) or revealing novel therapeutic opportunities (8/36).

Conclusion: Our findings demonstrate the feasibility and clinical value of centralised WGS for children with cancer. WGS offered additional clinical value, especially in diagnostic terms. However, our experience highlights the need for local expertise in scrutinising and clinically interpreting centrally derived variant calls for individual children.

Fig. 1. Bar chart indicating total variant counts in 36 childhood cancer genomes, and the clinical insight informed by these data.

The panels beneath the bar chart indicate the clinical insight to inform diagnosis (U: uninformative, C: consistent, R: refine, M: modify), prognosis (P: informative) and therapy (T: informative) in each case. The number of somatically acquired indels was greatly elevated in samples prepared using a PCR based (nano-prep) protocol (b) compared to sequencing libraries prepared without PCR (a). MB medulloblastoma, EP anaplastic ependymoma, PB pineoblastoma, LGG/HGG biphasic neuroepithelial tumour, PA pilocytic astrocytoma, PXA glioma with molecular features of pleomorphic xanthoastrocytoma, DLGNT diffuse leptomeningeal glioneuronal tumour, DNET dysembryoplastic neuroepithelial tumour, AB astroblastoma, ACC adrenocortical carcinoma, HB hepatoblastoma, NB neuroblastoma, G-NB ganglio-neuroblastoma, WT Wilms’ tumour, RCC renal cell carcinoma, RMS rhabdomyosarcoma, US undifferentiated sarcoma, ES Ewing’s sarcoma, OS osteosarcoma, CIFS congenital infantile fibrosarcoma, IT immature teratoma, OV_GRAN ovarian granulosa cell tumour, LYM high-grade B cell lymphoma.

Fig. 2. Detailed depiction of mutated cancer genes in 36 childhood cancer genomes, indicating the clinical outcome of each variant.

For each variant, information on diagnosis (U: uninformative, C: consistent, R: refine, M: modify) prognosis (P: informative), therapy (T: informative) and germline status (G: informative) is indicated. Variants that were previously known via standard of care are indicated with black borders, whereas those which are novel are indicated with red borders. For each mutated gene, a tally of known and novel variants is presented as filled red or black bars on the right of each row.

Fig. 3. Novel ZNF394-BRAF fusion gene identified in patient P2847 via a tandem duplication on chromosome 7.

The tumour specimen in this case was bi-phasic with both high- and low-grade components (a). The high-grade tissue was selected for whole-genome sequencing. The tandem duplication apposed exons 1–2 of ZNF394 with exons 10–18 of the BRAF oncogene preserving the BRAF kinase domain consistent with functional validity (b).