Prognostic Utility of Targeted Circulating Cell-Free DNA versus Formalin-Fixed Paraffin-Embedded DNA Mutation Analysis for Advanced Lung Cancer

Abstract

Background: Cancer burden is a globally growing problem. Early diagnosis and targeted treatment decrease patients’ death rate, pain, and treatment expenses. Liquid biopsy can be used for early cancer detection, treatment selection, and progression and treatment response monitoring. We evaluated the performance of circulating cell-free DNA (cfDNA) and formalin-fixed paraffin-embedded (FFPE) tissue DNA analyses using a commonly employed targeted therapeutic pathway in predicting the outcomes of patients with lung cancer, a common cancer with a generally poor prognosis. Methods: Patients with advanced lung cancer (n = 106) and predominantly adenocarcinoma histology donated blood samples at baseline and progression (n = 22), with matched archival FFPE biopsy samples being available for 75 patients. We set up a targeted 21-amplicon sequencing workflow for the analysis of mutations in nine genes (ALK, AKT1, BRAF, EGFR, ERBB2, KRAS, MET, NRAS, and PIK3CA). Molecular characteristics such as cfDNA concentration, number of mutations, and mutation occurrence in specific genes, were analyzed in respect of clinical outcome. To identify factors associated with overall survival (OS), a multivariate analysis was performed, using the technique of Cox regression. Results: The median age of the study population was 67 years with 53.8% male patients. The patients had primarily adenocarcinoma (79.2%) and stage IV disease (83%). The median OS was 327 (95% CI 275–514) days. Higher cfDNA concentrations were associated with poorer OS (HR = 1.670, 95% CI 1.108–2.516, P = 0.014). Detectable mutations [variant allele frequency (VAF) > 0.8%] were found in 63 (59%) patients’ baseline samples (median VAF = 1.1%). OS was significantly improved among patients with at least one mutation detected in cfDNA than among those with no mutation (HR = 0.477, 95% CI 0.313–0.727, P = 0.0006). In baseline cfDNA, patients with mutations of VAFs < 5% had significantly better OS compared to patients with mutations of VAFs ≥ 5% (HR = 3.510, 95% CI 1.672–7.370, P = 0.0009). Patients with slowly progressing disease had significantly more cfDNA mutations than did those with rapid cancer progression (P = 0.045). EGFR alterations explained half (16/33) of slowly progressing cases (median OS 953 days). All associations in FFPE biopsy material were statistically insignificant. *Corresponding author: Neeme Tõnisson, Estonian Genome Centre, Institute of Genomics, University of Tartu, Riia 23b, Tartu 51010, Estonia Check for updates ISSN: 2643-4563 DOI: 10.23937/2643-4563/1710033 Tamm et al. Int J Oncol Res 2022, 5:033 • Page 2 of 13 • Introduction The usefulness of the analysis of circulating cell-free DNA (cfDNA) from blood (“liquid biopsy”) for cancer diagnosis, monitoring, and treatment selection has been firmly established [1-4]. Liquid biopsy is an attractive alternative to more-invasive interventional solid-tissue biopsy for the guidance of therapeutic management based on somatic cancer variants [5]. cfDNA levels or variant allele frequencies (VAFs; fractions of cfDNA harboring a specific alteration) can be monitored longitudinally as potential prognostic biomarkers [6,7]. cfDNA mutation profiling by next-generation sequencing (NGS) has very high sensitivity [8,9]. For example, the incorporation of white blood cells to filter out somatic mutations associated with clonal hematopoiesis can enable the detection of mutations with VAFs below < 1% [10]. In addition, plasma cfDNA sequencing reduces the possibility that clinically relevant mutations in late-stage cancers are missed due to the issue of heterogeneity for biopsies obtained from single metastatic sites [11,12]. cfDNA analysis has been used successfully to identify actionable mutations in the epidermal growth factor receptor (EGFR) gene and other genes for the targeted treatment of non-small cell lung cancer (NSCLC) [1315]. Despite these potential advantages, however, the analysis of DNA from formalin-fixed paraffin-embedded solid-tumor tissue samples (FFPE DNA) remains much more common in the clinical setting. Several challenges associated with the use of cfDNA analysis, such as the accurate quantification and interpretation of VAFs, and the interpretation of variants with consideration of the complexity of the tumor mutational landscape remain [16,17]. Along with the lack of technical standardization, these challenges have led to some skepticism concerning the implementation of cfDNA analysis in daily clinical practice [17,18]. Emerging guidelines and standardized recommendations for cfDNA analysis and liquid biopsy in general are crucial for the development of this field [3,5,19,20]. This study was performed to evaluate the prognostic value of targeted cfDNA analysis relative to that of FFPE DNA analysis for advanced lung cancer. We hypothesized that differences in patients’ cancer progression and overall survival (OS) would be reflected in mutation profiles determined by both methods. We set up an NGS workflow to analyze cancer-related mutations in the nine frequently mutated genes in the EGFR pathway. Genetic findings from cfDNA and FFPE DNA analyses, as well as cfDNA concentrations, were compared with patients’ clinical profiles and therapeutic outcomes.