401P - The use of next generation sequencing (NGS) to guide patient selection for phase 1 clinical trials

Date 10 October 2016
Event ESMO 2016 Congress
Session Poster display
Topics Clinical Research
Basic Scientific Principles
Presenter Rowan Miller
Citation Annals of Oncology (2016) 27 (6): 114-135. 10.1093/annonc/mdw368
Authors R.E. Miller1, N.F. Brown1, A. Speirs2, H.M. Shaw1, S. Adeleke2, P. Gougis1, P. Bennett3, T. Meyer4, C. Swanton5, M. Forster4, R.S. Kristeleit4
  • 1Uclh/nihr Clinical Research Facility, University College London Hospital, W1T 7HA - London/GB
  • 2Medical Oncology, UCL - University College London, WC1E 6BT - London/GB
  • 3Sarah Cannon-ucl Laboratories, UCL- Advanced Diagnostics Molecular Profiling Laboratory,, WC1E 6JA - London/GB
  • 4Medical Oncology, University College London Cancer Institute, WC1E 6DD - London/GB
  • 5The Francis Crick Institute, UCL - University College London, London/GB



Therapeutically targeting actionable mutations in cancer may increase response rates in Phase I clinical trials. We undertook a pilot study to assess the feasibility and therapeutic benefit of incorporating NGS screening into the patient pathway for phase 1 cancer trials.


NGS tumour profiling was performed using a 22 gene amplicon-based panel (Life Technologies Colon & Lung V2) on 117 consecutive patients (pts) referred over a 13 month period for Phase I trials. BRCA1/2 analysis was performed in pts with epithelial ovarian cancer.


117 pts (67% female) with a median age of 59 (range 22-78) years were included. Common tumour types were ovarian (n = 20), colorectal (n = 16), breast (n = 13), endometrial (n = 12) and lung (n = 8) cancer. NGS was successfully performed in 108 (92%) pts with a median time to results of 12 days (range 6-39). 82% of pts (89/108) had a detected variant in ≥ 1 gene with an average of 3 variants (range 0-26) in 2 genes (0-10) per case. Common mutations included TP53 (69%), KRAS (14%), PIK3CA (11%) and SMAD4 (9%). BRCA1/2 mutations were present in 11 (55%) ovarian cancer pts. Overall, 49 (45%) pts had ≥ 1 actionable mutation. Detected variants were reviewed in a local genomics review board to assess actionability prior to considering therapy. 53 pts were commenced on a Phase I trial; 18 (34%) were genotype directed. Median duration on trial was 73 days for pts on genotype (7-260 days) or non-genotype (20-582) directed trials with 50% and 24% of allocated patients continuing on study respectively. Of pts evaluable for response (n = 47), partial response (PR), stable disease (SD) and progressive disease (PD) were observed in 50%, 29% and 21% of pts on genotype directed trials and 20%, 37% and 43% of pts on non-genotype directed respectively. Excluding pts on BRCA1/2 directed trials, PR, SD and PD were observed in 33%, 33% and 33% of pts respectively in genotype-directed studies.


NGS is feasible in real time and may affect clinical outcome in the phase 1 setting. Almost half of pts had a potentially actionable mutation. Initial response rates for pts treated on genotype-driven trials are encouraging. Benefit is likely to be augmented using a broader NGS panel which is planned for future assessment.

Clinical trial identification

Legal entity responsible for the study

UCLH/NIHR Clinical Research Facility


UCLH/NIHR Clinical Research Facility


P. Bennett: Employee of UCL- Advanced Diagnostics Molecular Profiling Laboratory, Sarah Cannon-UCL Laboratories. All other authors have declared no conflicts of interest.