Introduction. Immune-checkpoint inhibitors (ICIs) block inhibitory checkpoint molecules on T cells and antigen-presenting cells. ICIs have shifted treatment guidelines for multiple solid malignancies. Unfortunately, widespread clinical implementation of ICI treatment has been complicated by commonly occurring immune-related adverse events (irAEs) that involve diverse organs and currently represent a major challenge in cancer immunotherapy. This study was focused on identifying transcriptomic differences in ICI-treated cancer patients based on the development of irAEs.
Methods. This is a prospective multicentre cohort study (NCT04631731) recruiting cancer patients treated with ICIs at WSLHD. Peripheral blood was collected at baseline and in 6 weeks after first ICI treatment (post-cycle 2) as well as after the development of irAEs. Immune cells were isolated using Ficoll method and bulk RNA sequencing was done on Novaseq S200 cycle lane. Data transformation was completed in R with Tidyverse and edgeR packages. Functional annotation was performed using ConsensusPath-DB.
Results. N=71 patients were recruited to this pilot analysis and had paired blood samples taken. Patients were treated with: (A) single ICI, n=18 (B) Ipilimumab + Nivolumab, n=30 (C) ICI + chemotherapy, n=13 (D) ICIs + tyrosine kinase inhibitors, n=4 (E) ICIs + Bevacizumab, n=6. Within this cohort the top three malignancies were: (1) non-small cell lung cancer, n=18 (25.3%) (2) pleural mesothelioma, n=8 (11.2%) and (3) melanoma, n=7 (9.8%). N=25 (35.2%) had irAEs, including n=6 with grade (G) 3 hepatitis, n=1 G.4 hepatitis and n=8 G.2-3 colitis. The differential analysis established that 432 genes were significantly dysregulated in immune cells between (1) pre-treatment and (2) week 6 post-ICI commencement. Furthermore, the multivariate analysis adjusted by the type of therapy revealed that 102 genes were significantly upregulated only in patients from group B and were related to adenosine P1 receptors and oxidative stress. Finally, a pooled analysis of only patients with irAEs identified 160 significantly upregulated genes at toxicity onset as compared to their pre-treatment data. The functional annotation revealed that pathways related to Th17, Th2 and Th1 activation were enriched by those genes, although their role in ICI toxicity is yet to be defined.
Conclusions. The initial results revealed the presence of distinct differentially expressed genes in patients with irAEs at pre-treatment stage. Next, our research confirmed that combination of ICIs uniquely disrupts gene expression in immune cells. The ongoing recruitment and analysis will further explore the precise roles and functions of the identified genes and pathways.