ctDNA kinetics throughout first-line AI and palbociclib using a tumor-informed structural variant-based ctDNA assay: Retrospective analysis of PADA-1 samples.

L Luc Cabel V Veronique D'hondt (Institut du Cancer de Montpellier (ICM), Montpellier, France) T Thomas Bachelot P Patrick Soulie (ICO Paul Papin, Angers, France) C Chantal Bernard-Marty (Clinique Pasteur ONCOSUD, Toulouse, France) P Pascal Boucher (R&D Unicancer, French Breast Cancer Intergroup, Paris, France) C Clara Guyonneau (R&D Unicancer, French Breast Cancer Intergroup, Paris, France) A Anne Pradines (Claudius Regaud Institute, Toulouse, France) V Völundur Hafstad (SAGA Diagnostics, Morrisville, NC) C Céline Callens S Sofia Birkealv (SAGA Diagnostics, Lund, Sweden) M Miguel Alcaide Torres (SAGA Diagnostics, Morrisville, NC) J Jennifer Yen (SAGA Diagnostics, Morrisville, NC) K Karen Howarth (SAGA Diagnostics, Morrisville, NC) F Frederique Berger (Institut Curie, Biostatistics Unit, Paris, France) F François Clément Bidard

Abstract

3050 Background: With standard assays (Limit of Detection [LoD95] at or >100 PPM), most patients with ER+/HER2- mBC show ctDNA "clearance" within a few weeks after initiating first- line AI+CDK4/6i, with ctDNA becoming detectable again only at or shortly before progression. To gain a more comprehensive view of ctDNA kinetics in response to therapy, we analyzed serial plasma samples from PADA-1 using a tumor-informed assay with a 100-1,000-fold higher sensitivity (LoD95 <10PPM). Methods: PADA-1 allowed the collection of plasma at baseline and every two months during first-line AI+palbociclib (PAL). A total of 489 serial plasma samples from 45 patients in PADA-1 were analyzed using an ultrasensitive structural variant-based ctDNA assay. WGS was performed on tumor material from archival primary or metastatic biopsies, and personalized multiplex dPCR assays tracking up to 16 somatic structural variants were used for ctDNA monitoring. Results: At baseline, before AI+PAL start, ctDNA was detected in N=40/41 evaluable patients (98%). ctDNA remained detectable in 381/489 (78%) samples during therapy. ctDNA was persistently detectable and quantifiable (i.e. detected at all timepoints) throughout first-line therapy in most initially positive patients (24/40, 60%). With this ultra-sensitive assay, we report, for the first time, that ctDNA follows a recurrent kinetic pattern: (1) a steep decrease during the first 6 months of treatment, (2) a nadir reached at 6–12 months, and (3) a continuous increase until radiological disease progression (or discontinuation of follow-up). In some patients (15/40, 38%), ctDNA became undetectable at some point during therapy despite the high sensitivity of the assay, with sustained ctDNA clearance associated with no risk of impending tumor progression. Conclusions: Serial analyses with a highly sensitive ctDNA test uncover the molecular trajectory of tumor response to first line AI+PAL. The presence of an early ctDNA nadir, followed with a later slow and continuous increase of ctDNA, prior to radiological progression, highlights the dynamic nature of ER+ HER2- mBC response to therapy. This approach can complement imaging-based disease monitoring, extends the clinical utility of ctDNA monitoring beyond individual mutations, and opens new perspectives of therapeutic interventions and treatment adaptation. Clinical trial information: NCT03079011 .

Article Details

Volume / Issue Vol. 44, Issue 16_suppl
Published June 01, 2026
Pages 3050-3050
ISSN 0732-183X
Publisher Lippincott Williams & Wilkins

Journal Info

Journal of Clinical Oncology

Lippincott Williams & Wilkins

ISSN: 0732-183X Health Sciences

Authors (16)

L

Luc Cabel

V

Veronique D'hondt

Institut du Cancer de Montpellier (ICM), Montpellier, France

T

Thomas Bachelot

P

Patrick Soulie

ICO Paul Papin, Angers, France

C

Chantal Bernard-Marty

Clinique Pasteur ONCOSUD, Toulouse, France

P

Pascal Boucher

R&D Unicancer, French Breast Cancer Intergroup, Paris, France

C

Clara Guyonneau

R&D Unicancer, French Breast Cancer Intergroup, Paris, France

A

Anne Pradines

Claudius Regaud Institute, Toulouse, France

V

Völundur Hafstad

SAGA Diagnostics, Morrisville, NC

C

Céline Callens

S

Sofia Birkealv

SAGA Diagnostics, Lund, Sweden

M

Miguel Alcaide Torres

SAGA Diagnostics, Morrisville, NC

J

Jennifer Yen

SAGA Diagnostics, Morrisville, NC

K

Karen Howarth

SAGA Diagnostics, Morrisville, NC

F

Frederique Berger

Institut Curie, Biostatistics Unit, Paris, France

F

François Clément Bidard