CALIBRATION: Can early changes in circulating tumour DNA (ctDNA) predict durable tumour responses in patients with advanced Oesophageal Adenocarcinoma receiving anti PD-L1 antibody durvalumab?
Session type: E-poster/poster
Theme: Immunology and immunotherapy
Response of oesophageal adenocarcinomas (OAC) to immune-checkpoint blockade (ICB) is modest and clinicians lack biomarkers for patient (pt) selection. OAC's mutational signature profile suggest 'mutagenic' & 'DDR' subtypes might respond to ICB. Furthermore, ctDNA can detect treatment & genomic variations before other tests.
Prospective, open-label, pilot trial of durvalumab (1500mg/4week) for advanced OAC. Intensive blood sampling (PBMCs, plasma) & tumor biopsies are taken at pre-defined time-points (screening/week12/disease progression). Tumour-derived variant allele fractions (VAF) & copy-number alterations (CN) are analysed using Guardant360 assay. The primary objective is to determine if early changes (at week 4/7) in ctDNA (VAF, CN) correlates with long-term radiological responses (week26). Enrolment of 19 evaluable pts is planned. Translational studies will analyse the effect of durvalumab on the immuno-genomic landscape of OAC. We report the interim data.
7 pts were recruited (all male, median age: 71.7 yr, 1 patient HER2+), 6 received durvalumab. Prior chemotherapy: perioperative only in 2/6 pts, 4/6 had ≥1 prior treatment lines. Three pts received ≥ 3 doses of durvalumab. Adverse events (AE): 68% Grade 1-2, 32% (8/25) Grade 3 (none drug-related). Two drug-related AE (hyperamylasemia, hypothyroidism) led to dose delay and treatment withdrawal, respectively. Both resolved. Primary endpoint: changes in “total-ctDNA fraction" and RECIST1.1 response were evaluable in 4/6 pts:
Also, individual gene changes reflect RECIST responses:
-Pt3 had PR & ≥10% relative reduction in VAFs (PI3KCA, TP53, APC, ERBB2) & CNs (CDK4, FGFR2, CCNE1).
-Pt2 had PD & ≥10% relative increase of VAFs (TP53, BRCA1, ATM) & CN (KRAS, EGRF).
Durvalumab was well tolerated with potential activity in OAC. Data suggest a relationship between changes in ctDNA & RECIST response. Mature data is required. Recruitment & translational studies are ongoing, mutational signature profile (WGS data) of will be presented
Characterisation of multiple biomarkers throughout a patient's treatment, in the tumour bed and the periphery, will hep understand response to ICBs from a multidimensional perspective and can be used for treatment allocation.