1. Introduction
Generally, the term ”biomarker” refers to any biological indicator whose variations are related to the definition of the pathological condition, to the definition of prognosis, to the follow-up or to the definition of the results of the therapeutic response. The recent discovery of genetic alterations that cause, in a not neglected part of tumor patients, neoplastic disease encouraged the formulation of molecular target therapies that arise from the evaluation of precise molecular targets. In the context of predictive molecular pathology, liquid biopsy is a venous blood sampling on which molecular analyzes can be performed to search for clinically relevant mutations in the EGFR gene in patients with non-small cell lung cancer (NSCLC). This is usually performed when it is not possible to have a tissue sample representative of the neoplastic lesion at baseline or to monitor any resistance indicator or mutations acquired after a first line of treatment with tyrosine kinase inhibitors (TKI) .1 In particular, when it is not possible to perform the analysis molecular on tissue sample (difficult access to the lesion, comorbidity , especially in elderly patients, refusal by the patient, etc.) liquid biopsy can be a valid diagnostic solution. The liquid biopsy cannot replace the tissue biopsy because, to our knowledge, finding a certain mutation in the blood does not necessarily imply that the mutation is expressed at a structured clinically relevant frequency within the selected tumor (passenger mutation). Even the limited amount of circulating tumor DNA can be a boundary that can lead to false negative results.2
In the last five years, the liquid biopsy has been used for different purposes in oncology. Today the new challenge in lung cancer is represented by the implementation of some screening tests o, although this requires a phase of test harmonization and sensitivity assessment.3,4
The molecules that can be tested in a blood sample in the context of predictive molecular pathology are different: circulating tumor DNA (ctDNA) which represents a fraction of the total free DNA (cfDNA), cells released into circulation by the tumor through active mechanisms and passive and exosomes which also represent a precious source of nucleic acids of tumor origin.5 The cfDNA represents, to date, the specimen of greatest interest because of its immediate availability with respect to tumor DNA extracted from tissue samples; however due to the low concentration and the scarce bioavailability in blood torrent, new approaches are being optimized in order to recover tumor nucleic acids starting from the lysosomal cargo.2,6 Clinical response data have shown that in addition to the sensitivity mutations, there is a ’wide range of different mutations that fall within exon 18, 19, 20 and 21 leading the patient to benefit from the target therapy.8
Thus they have a predictive potential and for this reason, new-generation gene sequencing (NGS) represents a valid analytical platform for the analysis of the mutations present in the cfDNA.9 However, the analysis spectrum for NSCLC patients does not only exhaust the EGFR gene; the search for oncogenic activating genomic translocations such as ALK, ROS, RET has become clinical practice, a routine that makes the patient eligible for treatment with TKIs starting from tissue samples.10This possibility represents the direction towards which scientific research is moving to optimize patient clinical management.11
Tumors never reach complete molecular stability and this issue means that, over time, they can develop resistance to target therapies.12, so it is a crucial decision-making step to identify what it is happening in the tumor. The presented study is aimed at assessing the diagnostic and predictive potential of liquid biopsy into management of lung cancer path of care.
2. Methods
From 2014-2019 liquid biopsy specimens were analyzed for EGFR sensitive and resistance mutations in Predictive Molecular Pathology (Department of public Health, University of Naples Federico II). For each patient 5 ml of plasma and serum sample were withdrawn and collected in Vacutainer tubes (BD, Plymouth, UK) by a dedicated nurse. Plasma and serum were immediately separated by two centrifugation steps at 2300 r.p.m. for 10 min, then supernatant (where nucleic acids were carried out) was stored at -20C® . Cell-free DNA (cfDNA) was purified by using the QIAsymphony robot (Qiagen, Hilden, Germany ) and the QIAsymphony DSPVirus/ Pathogen Midi Kit, according to the manufacturer’s instructions, finally cfDNA was eluted in a final volume of 35 µl ofnot DEPC water (Ambion). Libraries were produced on the Ion Chef (Thermofisher, Waltham, Massachusetts ) by using DL-8 kit and serum and plasma of 4 patients were processed for a total of eight specimens for run. After library amplification step performed by 22 cycles for cfDNA amplification and 6 cycles for library reamplification after barcoding, purified libraries derived from eight cfDNA samples were diluted to 60 pM and combined with eight additional cfDNA-derived libraries to obtain a 16 Ion Code pooled library. The two-pooled libraries were re-loaded into the Ion Chef instrument, and templates were prepared using the Ion S5 520-530 chef kit ( Thermofisher,Waltham, Massachusetts ). Finally, templates were loaded into the 520 chip and sequenced on S5 platform. Signal processing and base calling were evaluated by adopting the default base-caller parameters on Torrent Suite [v.5.0.2]. Results were visually inspected with the Golden Helix Genome Browser v.3.0. (Bozeman,MT, USA). Statistical analysis were carried out by adopting a binary logistic regression model on SPSS and IBM software system. Age and sex parameters were added as covariates.
Sample related to 5 years of activity regarding the molecular diagnostics performed with liquid biopsy in the predictive diagnostic laboratory of the AOU Federico II were analyzed. These patients came from the oncological network pathways related to lung cancer and refer to the HUB center of the AOU Federico II. The samples of 900 patients enrolled in the pathway were collected. The availability and epidemiological interest was verified.
After database review, 515 plasma or blood samples were included in the analysis (Fig 1). Descriptive analyzes were performed (Tab1). To verify the differences with respect to the report of liquid biopsy between patients with basal or progressing tumors, non-parametric analyzes were performed for 2 independent samples. Furthermore, the effect of the presence of mutations on the probability of tumor progression, taking into account age and sex, was evaluated through a multivariate logistic regression model (Tab2)