Tumour Immune Characterization in Adenoid Cystic Carcinoma Identifies Prognostic and Immunotherapeutically Relevant Messenger RNA Signatures.

Shengjin Dou,Di Chen,Rongrong Li,Yining Yang,Ning He,Ge Jin,Jiang Li,Guopei Zhu
DOI: https://doi.org/10.1200/jco.2020.38.15_suppl.e15191
IF: 45.3
2020-01-01
Journal of Clinical Oncology
Abstract:e15191 Background: Here, we profiled genetic hallmarks of Chinese adenoid cystic carcinoma (ACC) patients and immune cells infiltration levels by gene expression-based computational methods to investigate the correlation with immunotherapy. Methods: 54 patients with pathologically confirmed ACC admitted from January 2016 to October 2019 were included. Mutation and expression profiles were obtained by whole-exome sequencing (WES) and RNA-seq, respectively. A nomogram that integrated clinicopathological features with the immune signature to predict recurrence or distant metastasis probability was constructed by multivariate Cox regression. Results: As expected, we observed MYB fusion and somatic mutations in our cohort (57%, 31/54), strengthening the role of these aberrations as critical events in ACC. Mutations in NOTCH1 were also found in our cohort (19%, 10/54). Median tumour mutational burden (TMB) of these patients is 0.85 Muts/Mbp (0-11 Muts/Mbp). Analysis of WES data showed that all patients were microsatellite instability (MSI)-negative, and 17 of them were confirmed by conventional MSI-PCR testing. Lack of PD-L1 expression is 96.0% (n = 46/48) at a tumour proportion score 0% and 22.9% (11/48) of cases showed more than 10% CD8+ prevalence. We compared cancer types using an immune infiltration score (IIS) and an antigen presenting machinery expression (APM) score and find that ACC is among the lowest for both scores. Statistically significant correlations were found between IIS and distant metastasis-free survival (univariate analysis, HR = 2.64, 95% confidence interval [CI], 1.134 to 6.145, p = 0.023). Patients with higher IIS had shorter distant metastasis-free survival time (log-rank p = 0.020), and patients with higher T cell infiltration score (TIS) had shorter locoregional recurrence-free survival time (log-rank p = 0.050). In addition, we investigated the pretreatment immune profile of five ACC patients treated with PD-1 checkpoint inhibitors, and found that both TIS and APM were elevated in responding patient (with a partial response to camrelizumab) whereas they were in the lowest for patients with progressive disease on camrelizumab or pembrolizumab. Conclusions: Our study highlights the immune infiltration patterns and unravels mRNA signatures with prognostic utility and immunotherapeutic biomarker potential in ACC.
What problem does this paper attempt to address?