Abstract 71P
Background
Tusamitamab ravtansine is an antibody-drug conjugate of a humanized carcinoembryonic antigen (CEA)-related cell adhesion molecule 5 (CEACAM5)-specific monoclonal antibody linked to DM4. A phase I/II study showed tusamitamab ravtansine antitumor activity in pretreated patients (pts) with advanced nonsquamous NSCLC and high CEACAM5 expression. Here, we explore biomarker associations with tumor CEACAM5 expression by immunohistochemistry (IHC), and whether biomarkers predict objective response rate (ORR).
Methods
We assessed CEACAM5 expression by IHC, RNA sequencing, and whole exome sequencing (WES) on latest archival tumor samples; and circulating CEACAM5 (cCEACAM5) and CEA (cCEA). We enrolled 2 cohorts of pts with IHC CEACAM5 membrane expression at ≥2+ intensity: in ≥50% of tumor cells (high expressors, HEs, n = 64); and in ≥1% to <50% of tumor cells (moderate expressors, MEs, n = 28). Pts received tusamitamab ravtansine 100 mg/m2 IV every 2 weeks.
Results
cCEA and cCEACAM5 were strongly associated (Spearman rho, 0.9), with weak associations between IHC CEACAM5 and cCEA or cCEACAM5 (Spearman rho, 0.3 and 0.4, respectively). Higher levels of CEACAM5 mRNA were observed in CEACAM5 HEs vs MEs (P = 0.0027). EGFR and KRAS genetic alterations by WES were present in 44.8% and 65.5% of CEACAM5 HEs, respectively, and 21.4% and 78.6% of CEACAM5 MEs, respectively. Confirmed partial responses were seen in 13/64 HEs (ORR 20.3%) and 2/28 MEs (ORR 7.1%). In CEACAM5 HEs with available baseline (BL) cCEA data, 25/62 (40.3%) had a cCEA level ≥100 µg/L, with a median value of 71.6 µg/L (range 1–8809); corresponding values in CEACAM5 MEs were 7/28 (25.0%) and 12.4 µg/L (range 0.5–684). In response evaluable CEACAM5 HEs with available BL cCEA data (n = 61), ORR was 10/24 (41.7%) in pts with high cCEA (≥100 µg/L) and 3/37 (8.1%) in pts with low cCEA (<100 µg/L); corresponding ORRs in CEACAM5 MEs were 0/7 and 2/21 (9.5%).
Conclusions
In CEACAM5 HEs, high cCEA was associated with numerically greater ORR vs low cCEA (41.7% vs 8.1%). Associations were also observed between: cCEA and cCEACAM5; IHC CEACAM5, cCEA, and cCEACAM5; and IHC CEACAM5 and CEACAM5 tumor mRNA levels, but not between IHC CEACAM5 and actionable oncogenic drivers.
Clinical trial identification
NCT02187848.
Editorial acknowledgement
Medical writing was provided by Michael Stillman and Julian Martins, inScience Communications and was funded by Sanofi.
Legal entity responsible for the study
Sanofi.
Funding
Sanofi.
Disclosure
J.S. Lee: Financial Interests, Personal, Full or part-time Employment: Sanofi; Financial Interests, Personal, Stocks/Shares: Sanofi. E. Wang: Financial Interests, Personal, Full or part-time Employment: Sanofi. N. Ternès: Financial Interests, Personal, Full or part-time Employment: Sanofi. H. Wang: Financial Interests, Personal, Full or part-time Employment: Sanofi. E. Boitier: Financial Interests, Personal, Full or part-time Employment: Sanofi. A. Lartigau: Financial Interests, Personal, Full or part-time Employment: Sanofi. M. Chadjaa: Financial Interests, Personal, Full or part-time Employment: Sanofi. C. Dib: Financial Interests, Personal, Full or part-time Employment: Sanofi; Financial Interests, Personal, Stocks/Shares: Sanofi. G. Muzard: Financial Interests, Personal, Full or part-time Employment: Sanofi. S. Valence: Financial Interests, Personal, Full or part-time Employment: Sanofi. A. Remaury: Financial Interests, Personal, Full or part-time Employment: Sanofi. C.C. Palu: Financial Interests, Personal, Full or part-time Employment: Sanofi; Financial Interests, Personal, Stocks/Shares: Sanofi. A. Bauchet: Financial Interests, Personal, Full or part-time Employment: Sanofi; Financial Interests, Personal, Stocks/Shares: Sanofi. All other authors have declared no conflicts of interest.