Genetic heterogeneity and predictive biomarker for pulmonary sarcomatoid carcinomas

Highlights

• PSC containing carcinomatous and sarcomatous components had a mild heterogeneity from the genomic profiling.

• The two components may evolve from common ancestral cells.

• There may be relatively more opportunities for targeted therapy.

• Strong PD-L1 expression suggests that immunotherapy could be used as a potential therapy for PSC patients.

• Negative immune-responsive genes need to be screened out.

Abstract

Purpose

The aim of this study is to investigate the genetic heterogeneity (carcinomatous vs. sarcomatous components) and predictive biomarkers in patients with pulmonary sarcomatoid carcinoma (PSC).

Methods

Genetic alterations and biomarkers of immunotherapy were performed in a discovery set (n = 6) of PSC. Next-generation sequencing (NGS) on a pan-cancer gene panel was applied to detect the genetic alterations in each component, and the respective mutation profiling and tumor mutation burden (TMB) were compared as well. Immunohistochemistry (IHC) assay with SP263 antibody was used to detect the protein expression of programmed death-ligand 1 (PD-L1) in each component.

Results

Comparative genetic analysis revealed that the separate carcinomatous and sarcomatous components shared strikingly common mutations. TP53 (4/6, 66.7%) was the most common genetic alteration in 6 PSC patients. MET exon 14 skipping was detected in one case, accounting for 16.7%. An EZR-ROS1 fusion (EZR: intron10-ROS1: intron32) was identified in one case. The TMB of the two components was similar. Nevertheless, significantly higher PD-L1 expression was found in carcinomatous components compared to sarcomatous components. MDM2 amplification was detected in 2/6 (33.3%) of cases and STK11 mutation in 1/6 (16.7%) of cases.

Conclusions

PSC containing carcinomatous and sarcomatous components had a mild heterogeneity; the two components may evolve from common ancestral cells. High PD-L1 expression suggests that immunotherapy could be used as a potential therapy for PSC patients, while patients with negative immune-responsive genes need to be screened out. Altogether, these findings further highlight that the detection of genetic alteration and PD-L1 expression plays an important role in treatment of patients with PSC.

Introduction

Pulmonary sarcomatoid carcinoma (PSC) is a rare subtype of non-small cell lung cancer (NSCLC), accounting for about 0.4% of all lung malignancies [1]. PSC is defined as poorly differentiated NSCLC which contains a component of sarcoma or sarcoma-like element [2]. Histologically, PSC is manifested as biphasic malignancies, combining the carcinoma component and sarcomatoid component, sometimes with heterologous sarcomatous tissue. Hence, PSC was classified as pathologically heterogeneous group. Both epithelial and mesenchymal elements of tumor could be shown using immunohistochemistry (IHC). According to the most recent 2015 World Health Organization (WHO) classification, five subtypes of PSCs have been identified as follows: pleomorphic carcinoma, spindle cell carcinoma, giant cell carcinoma, carcinosarcoma and pulmonary blastoma[3]. PSC patients are often found in heavy smokers, with an elderly male predominance [1]. Compared to common NSCLC, PSC has a worse prognosis, higher resistance to chemotherapy and radiotherapy, and shorter disease-free survival (DFS) after surgical resection [4]. The median overall survival (OS) and 5-year survival rate of PSC are 7 months and 19.5%, respectively [5]. However, the main causes leading to these events remain unclear. Besides, it is still unclear whether tumor heterogeneity might affect the recurrence and treatment-resistance of PSC.

Recently, genomic profiling has been applied to investigate new strategies for improving the prognosis in patients with PSC [6,7]. Tyrosine kinase inhibitor (TKI) has been recommended for the treatment of NSCLC patients with epidermal growth factor receptor (EGFR) mutation or fusions involving anaplastic lymphoma kinase (ALK). Moreover, a novel and promising immune checkpoint inhibitor targeting the PD-1/PD-L1 axis has been recommended as first-line therapy for those with wild-type EGFR and ALK while overexpressing PD-L1 [8]. However, new therapeutic targets and biomarkers of immunotherapy response have not been well investigated for PSC. Previous studies reported the MET mutation leading to exon 14 skipping in 8 (22%) out of 36 PSC cases [9,10]. Moreover, deep sequencing analysis revealed that KRAS mutation is a potential biomarker of poor prognosis in PSC patients [11,12].

Immunotherapy offers a new therapeutic option for sarcoma patients, but its therapeutic effect needs to be further investigated in clinical trials [13]. Therefore, through the identification of genetic profiling using NGS technology, followed by detection of PD-L1 expression, treatment for tumor subsets can be facilitated by personalizing therapies.

In this study, carcinomatous and sarcomatous components were identified in 6 PSC patients who underwent surgery. Using a customized panel covering exons of 425 cancer-relevant genes and selected introns of 32 fusions genes, NGS was performed to interrogate genetic profiling. In order to provide evidence for the clonal origin of the carcinomatous, sarcomatous components and alternative therapeutic targets for immunotherapy, genetic alterations and tumor mutation burden (TMB) were evaluated. In addition, PD-L1 expression was assessed by IHC, SP263 antibody was used to identify whether PSC patients may benefit from anti-PD-1/PD-L1 immunotherapy.