1469PCirculating tumour DNA (ctDNA) analysis in patients (pts) with non-small cell lung cancer (NSCLC) treated with telisotuzumab vedotin (teliso-v), an antibody-drug conjugate targeting c-Met

Abstract Background ctDNA allows monitoring of response to therapy and serves as a liquid biopsy for pt selection for targeted therapies. We used samples from pts with NSCLC enrolled in an ongoing phase I/Ib trial (NCT02099058) treated with telisotuzumab vedotin (ABBV-399; teliso-v) alone and in com...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Annals of oncology 2019-10, Vol.30 (Supplement_5)
Hauptverfasser: Heist, R S, Motwani, M, Naumovski, L, Wu, J, Bach, B A, Lu, X, Kelly, K
Format: Artikel
Sprache:eng
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Abstract Background ctDNA allows monitoring of response to therapy and serves as a liquid biopsy for pt selection for targeted therapies. We used samples from pts with NSCLC enrolled in an ongoing phase I/Ib trial (NCT02099058) treated with telisotuzumab vedotin (ABBV-399; teliso-v) alone and in combination with erlotinib or nivolumab to assess the frequency of mutant alleles in ctDNA and the association between ctDNA levels and OR to teliso-v. Methods Plasma was collected pre- (C1D1) and post-treatment (≥C2D1) and analyzed using 64-gene PlasmaSELECT™ assay (PGDx). Correlations between ctDNA detection probability and best OR were assessed by Fisher’s exact test; if C2D1 samples were not available, C3D1 (if CR/PR/SD) or FV samples (if PD) were used. The sum of variant allele fraction for all somatic mutations (SumVAF) at C1D1 and C2D1 was compared between response groups and between matched C1D1 and C2D1 samples. Results 109 pts were analyzed (C1D1 [n = 108]). Among 56 pts without variant alleles detected at C1D1, ctDNA detection probability at ≥C2D1 was similar regardless of response (P = 0.51). For 52 pts with ctDNA detectable at C1D1, ctDNA detection probability at ≥C2D1 was higher for SD/PD (85%) vs CR/PR (55%), but not significant (P = 0.15). For all response groups SumVAF decreased from C1D1 to C2D1, suggesting a pharmacodynamic effect of teliso-v. SumVAF was significantly lower in pts with CR/PR vs SD/PD at C1D1 (P = 0.015) and marginally lower at C2D1 (P = 0.097). Table:1469P Pre- and post-treatment mean (stdev) sum of variant allele fraction by response categoryCR/PR n = 24SD n = 58PD n = 27P§ (FC) of SumVAF: SD/PD vs CR/PRPre-treatment (C1D1) n = 108*3% (6%)8% (15%)6% (15%)0.015 (2.88)Post-treatment (C2D1) n = 83†1% (5%)4% (7%)3% (9%)0.097 (2.87)P‡, C1D1 vs C2D1 Mean decrease of SumVAF from C1D1 to C2D10.01 2%0.04 3%0.02 4%*C1D1 data, missing for 1 patient with SD.†C2D1 data, missing for 26 patients (4 CR/PR, 11 SD, and 11 PD).‡By pair-wise t-test on 82 patients with both C1D1 and C2D1 data available.§By 2-group t-test.C, cycle; CR, complete response; D, day; FC, fold change; PD, progressive disease; PR, partial response; SD, stable disease; stdev, standard deviation; SumVAF, sum of variant allele fraction for all somatic mutations. Conclusions Baseline SumVAF levels and ctDNA detection probability at ≥C2D1 were lower with CR/PR vs SD/PD in pts with detectable ctDNA at C1D1. SumVAF decreased by C2D1 regardless of response, with more consistent red
ISSN:0923-7534
1569-8041
DOI:10.1093/annonc/mdz259.012