Circulating tumour DNA (ctDNA) as a tool to assess response and guide therapy adaptation in rectal cancer

Khakoo, Salim, Carter, P., Valeri, N., Shaikh, R., Jones, T., Begum, R., Rana, I, Picchia, S., Bali, M., Brown, G, Wotherspoon, A, Terlizzo, M., Von Loga, K., Ahmed, I., Watkins, D., Chau, I., Starling, N., Tait, D., Hubank, M. and Cunningham, D. (2018) Circulating tumour DNA (ctDNA) as a tool to assess response and guide therapy adaptation in rectal cancer. Annals of Oncology, 29 (Supplement 8), [1867P]. (doi:10.1093/annonc/mdy303.037).

Abstract

Background: Neo-adjuvant chemoradiation (CRT) is associated with varied response in patients (pts) with localised rectal cancer. Early identification of poor responders and pts at risk of developing systemic disease using ctDNA would allow tailoring of treatment.

Methods: Tissue and serial plasma was collected from 47 pts with localised rectal cancer scheduled to undergo long course CRT. Cell-free DNA (cfDNA) was purified from a mean of 3.6 ml plasma per time-point. Somatic variants were identified in tissue by sequencing using a targeted capture panel. Up to 3 variants per patient in genes of interest were tracked in plasma using custom TaqMan assays on a droplet digital PCR platform. Tumour response assessments were conducted according to RECIST. Statistical analysis included Fisher’s exact test and Spearman’s correlation.

Results: 62% of pts were male, median age 59, range 30-83. On baseline MRI, circumferential resection margin was involved or threatened in 75% and EMVI positive in 81%. Plasma was collected at a median of: 6 days (d) prior to CRT (baseline), 21 d from the start of CRT (mid) and 37 d after completion of CRT (end). The frequency (%) of mutation detection in tissue was: TP53 (85), APC (74), KRAS (36), PIK3CA (15), NRAS (4) and BRAF (2). ctDNA was detectable in 74% of pts at baseline and in 21% at mid and end of CRT. ctDNA detection increased with stage at baseline: stage 1 (n = 0/1), stage 2 (n = 3/5, 60%) and stage 3 (n = 32/41, 78%). Stage had no impact on detection at mid or end of CRT. At baseline, ctDNA was detectable in all 15 CEA secretors (≥5 ug/l) compared to 63% in 30 non-secretors (P = 0.008). Baseline ctDNA levels were not associated with ki67 tumour assessment. MRI response assessment of the primary tumour was not associated with ctDNA detection at any timepoint. 11 patients developed metastases of which 3 occurred after surgery. End of CRT ctDNA detection was higher in pts that developed metastases (64%) compared to those that did not (8.3% P = 0.0005). Detection of ctDNA at baseline that persisted at mid CRT was also higher in pts that developed metastases (36% vs 11%, p = 0.07).

Conclusions: ctDNA detection can help identify rectal cancer pts with localised disease at risk of developing metastases. These pts could benefit from earlier intervention with systemic therapy.

This record has no associated files available for download.

Contributors

Author: Salim Khakoo

Download statistics