Advice

Clinical and technical evidence

Clinical and technical evidence

For this technology, a variation from the interim process and methods statement was made. No literature search was done. Instead, experts and relevant charities provided the most relevant or best available published evidence relating to the clinical effectiveness of the technology. Further information about how the evidence for this briefing was selected is available on request by contacting mibs@nice.org.uk.

Published evidence

This briefing summarises 1 analytical validity and diagnostic accuracy study of 132 samples. This study was done at The Royal Marsden NHS Foundation Trust using clinical samples obtained from laboratories globally. Reported outcomes were:

  • limit of detection

  • precision

  • panel sensitivity

  • range of variant allele frequency detected

  • performance and variant detection comparison in paired fresh frozen and FFPE clinical DNA samples

  • detection of known variants

  • detection of rearrangements.

The study concluded that the NGS panel detected 94 of 95 (98.9%) well-characterised genetic abnormalities in 33 clinical specimens and 13 cell lines (including single nucleotide variants, insertions or deletions, amplifications, rearrangements and chromosome losses).

Table 2 summarises the clinical evidence as well as its strengths and limitations.

Overall assessment of the evidence

The technology is still in development, so the published evidence base is limited in terms of quantity. Evidence on the analytical validity and diagnostic accuracy of the first version of the test (which included 78 genes) is available, and a number of clinical trials are ongoing or planned. The current version, which includes 92 genes, has been verified but the results were not publically available at the time of preparing this briefing.

Table 2 Summary of selected studies

Izquierdo et al. (2017)

Study size, design and location

Analytical validity and diagnostic accuracy study of 132 samples from laboratories worldwide. London, UK.

Intervention and comparator(s)

Intervention: in-house, NGS panel comprising 78 genes involved in childhood cancers. Samples included: Horizon Discovery cell blends (n=4), childhood cancer cell lines (n=15), FFPE clinical samples (n=83) and FF clinical samples (n=30).

Comparators: cell blends containing known SNVs (n=528) and small indels (n=108) were used to determine panel sensitivities of ≥98% for SNVs and ≥83% for indels, and panel specificity of ≥98% for SNVs.

Key outcomes

Panel sensitivity

All background SNVs and cancer-specific SNVs were detected, resulting in a sensitivity of over ≥98%. 90 of 108 background indels and 15 of 17 cancer-specific indels were detected, resulting in a sensitivity of ≥83%.

Panel specificity

The specificity of cancer-specific SNVs was ≥98%. PPV was ≥98% and NPV was ≥98%. There were insufficient true-negative indels (n=3) to determine specificity for indels.

Detection of known variants

100% of the 90 known genetic variants in 41 samples (in 13 cell lines, 14 FFPE and 14 FF samples) were detected.

Detection of rearrangements

94 of 95 (98.9%) well-characterised genetic abnormalities in 33 clinical specimens and 13 cell lines (including SNVs, indels, amplifications, rearrangements and chromosome losses) were detected by the panel.

Strengths and limitations

The study is written up transparently, with supplementary data available for independent scrutiny. Despite small numbers and low prevalence of these cancers in children, the use of paediatric clinical samples from an international collaboration should yield more generalisable findings than a smaller, single-centre study design.

However, as of December 2017, the work was still undergoing peer review at the OncoTarget journal.

The authors note that validation of the panel for the detection of rearrangements (translocations) was limited by small sample size (5 FFPE sarcoma samples), so further work is in progress at this stage.

Abbreviations: FF, fresh frozen; FFPE, formalin-fixed paraffin embedded; indels, insertions or deletions; NGS, next-generation sequencing; NPV, negative predictive value; PPV, positive predictive value; SNV, single nucleotide variants.

Recent and ongoing studies

The NGS panel is currently being used for the following studies:

  • BEACON: a randomised phase IIb trial of bevacizumab for refractory and relapsed neuroblastoma in children. EurodraCT identifier: 2012-000072-42. Status: ongoing.

  • CRISP: a phase IB of crizotinib for ALK-, ROS1- or MET‑positive malignancies in children. EurodraCT identifier: 2015-005437-53. Status: ongoing.

  • BIOMEDE (Biological Medicine for Diffuse Intrinsic Pontine Glioma Eradication). ClinicalTrials.gov identifier: NCT02233049. Status: recruiting.

  • eSMART (European Proof of Concept Therapeutic Stratification Trial of Molecular Anomalies in Relapsed or Refractory Tumours). ClinicalTrials.gov identifier: NCT02813135. Status: recruiting.

  • FaR-RMS (Frontline and Relapse study in Rhabdomyosarcoma). Status: awaiting full approval.

    The NGS panel forms the genetic testing basis for an ongoing study at the Royal Marsden Hospital (NGS sequencing CCR4924) and the national CCLG METEOR trial. It is also being used in the Children with Cancer UK‑funded IMPACT trial and as part of the Cancer Research UK Stratified Medicine Programme 2.