1485P - Response prediction using NanoString nCounter technology and NGS panel sequencing in neoadjuvant chemotherapy in patients with esophagogastric adenocarcinoma

Background

Bachground Perioperative radio-/chemotherapy is accepted standard therapy in patients with locally advanced adenocarcinoma of the esophagogastric junction or stomach. However, 30-85% of patients do not respond to this treatment. Aim of our study was the identification of predicting molecular markers in pretherapeutic endoscopic tumor biopsies from patients with response (regression score 1 according to Becker) versus non-response (Becker2/3) to preoperative chemotherapy.

Methods

Out of 626 screened patients, FFPE biopsies from 36 patients (Becker1 n=11, Becker2 n=9, Becker3 n=16) were selected for RNA expression analysis using the NanoString nCounter System (Cancer Pathway and Immuno Profiling Panel) and RNA/ DNA expression analysis with TruSight^TM Tumor 170. Consecutively, a differential expression analysis was conducted for detection of molecular divergencies. Panel sequencing using Qiagen Lung V3 Panel and TruSight^TM Tumor 170 (Illumina) were performed to detect genomic abnormalities.

Results

Heatmap and principle component analysis showed no clustering according to regression grades. ELK1 und ERRB2 were significantly overexpressed in Becker1 compared to Becker2/3. This was however due to single overexpressing samples. Furthermore, no differences were seen in expression of immune checkpoint inhibitors (e.g. PD-L1 or CTLA4) as well as cell type profiling of infiltrating immunologic active cells. NGS sequencing using Qiagen Lung V3 Panel indicated among Becker2/3 samples more often single nucleotide variations (SNV: 76%) and copy number variations (CNV: 61%) compared to Becker1 samples (SNV: 44%; CNV: 33%).

Conclusions

In our study, we could not identify any predictive RNA expression pattern in pretherapeutic biopsies. However, mutational analysis showed increased numbers of mutations (SNV and CNV) in non-responders in comparison to responders. TruSight^TM Tumor 170 data will be presented at the meeting. Funded by Berlin Institute of Health_ PRO_383.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

Severin Daum.

Funding

Berlin Institute of Health.

Disclosure

All authors have declared no conflicts of interest.