Skip to main content
. 2019 Mar 12;30(6):945–952. doi: 10.1093/annonc/mdz085

Figure 2.

Figure 2.

Identification of optimal circulating tumour DNA (ctDNA) early timepoint for prediction of progression-free survival (PFS) length in the development cohort. (A) Circulating tumour DNA ratios (CDRs) at designated timepoints in the first 4 weeks of study treatment, separated by short and long PFS in study part A. Differing numbers in the long and short groups reflect missed sample collection timepoints in some patients. Long and short PFS was determined by a scan at 12 weeks on study, the time-point of the first scan in part A. C2D1 (CDR28) was the strongest predictive timepoint P = 0.0007, P value Mann–Whitney U test. (B) PFS for development part A patients split by CDR28 suppressed (CDR28<0.25) versus CDR28 high. P value log rank test. (C) Longitudinal tracking of a PIK3CA c.1624G>A (p.E542K) mutation in a patient classified as a long PFS by ctDNA demonstrating successful suppression of ctDNA before rise before progression. (D) Longitudinal tracking of a TP53 c.815T>C (p.V272A) mutation in a patient classified as a short PFS by ctDNA demonstrating failure to suppress ctDNA in the first 4 weeks of treatment.