Pro-inflammatory tumour profile predicts complete pathological response to neoadjuvant chemoimmunotherapy

In the phase 2 NADIM trial, upregulated expression of genes involved in type II interferon signalling as well as the presence of M1 macrophages, follicular helper T cells, activated NK cells, and resting dendritic cells in the tumour predict complete pathological response to neoadjuvant chemoimmunotherapy in patients with stage IIIA non-small cell lung cancer (NSCLC).

Neoadjuvant chemoimmunotherapy is increasingly explored as a treatment option for patients with NSCLC. Recently, the open-label, single-arm, phase 2 NADIM trial (NCT03081689) demonstrated improved outcomes with nivolumab plus paclitaxel/carboplatin chemotherapy for patients with stage IIIA NSCLC; 82.9% of participants achieved a major pathological response (>90% necrosis), and 63.4% of participants achieved a complete pathological response (100% necrosis) [1]. In a post-hoc analysis, gene expression profiles were determined to identify biomarkers that might predict complete pathological response. Dr Marta Casarrubios (Hospital Puerta de Hierro-Majadahonda, Spain) presented the results [2].

Pre-treatment tumour samples of 14 participants from NADIM were subjected to RNA sequencing, as well as immune cell subtyping. Results were correlated with pathological response groups (9 complete and 5 non-complete responders). Non-complete responders had either a major response or an incomplete response (<90% necrosis ).

A total of 25 genes appeared to be differentially expressed between complete and non-complete responders. Of these, 17 were upregulated in complete responders compared with non-complete responders; 8 genes were downregulated in complete responders. Upregulated genes included genes involved in type II interferon signalling (FASLG, CXCL13, CXCL10, CXCL9, and IFNG), natural killer (NK) cell markers (NCR1, KIR2DL3, GNLY), genes involved in lymphocyte enrichment and activation (NKG7, GZMB), and 3 tumour antigens (GAGE1, MAGEA10, MAGEA1). Together, these genes depict a pro-inflammatory phenotype in pre-treatment tumour samples of complete pathological responders. Upregulation of IFNG (interferon gamma), GZMB (granzyme B), and NKG7 (NK cell granule protein 7) was associated with high sensitivity and specificity for complete pathological response. Area under the receiver operating characteristic curve (AUROC) for response prediction based on IFNG and GZMB expression was 1.000 (P=0.0027) and 0.911 (P=0.0136), respectively. Downregulation in complete responders was observed in genes involved in pathways related to tumour markers, proliferation, and PD-1 signalling.

Finally, immune cell subtyping demonstrated that a higher proportion of follicular helper T cells, activated NK cells, M1 macrophages, resting dendritic cells, and an absolute score of immune cells were present in tumour samples of complete responders. Frequency of M1 macrophages was significantly associated with complete pathological response. The AUROC for response prediction based on M1 macrophages proportion was 0.9778 (P=0.004).

“A pro-inflammatory profile measured in the pre-treatment tumour tissue by RNA sequencing is associated with complete pathological response following neoadjuvant chemoimmunotherapy in patients with stage IIIA NSCLC,” summarised Dr Casarrubios.

1. Provencio M, et al. Lancet Oncol. 2020; 21: 1413–1422.


2. Casarrubios M, et al. Immune gene signatures for predicting pathological response of NSCLC patients treated with neoadjuvant chemoimmunotherapy. Abstract 157MO. ELCC 2022 Virtual Meeting, 30 March–02 April.

 

Copyright ©2022 Medicom Medical Publishers





Posted on 25 May, 2022