Page 73 - GTM-1-1

Global Translational Medicine ctDNA in management of patients with NSCLC

of mortality from lung cancer by 20% in a 2-year follow-up early detection indicator . Cohen et al. integrated the
[12]
visit . Similar small-scale investigations emerged in mutation information in cfDNA and traditional protein
[37]
succession, but none of them yielded findings of comparable biomarkers to elevate the efficiency of early-stage cancer
significance or reliability of NLST trial outcomes until the evaluation and invented Cancer-SEEK method, which
Dutch-Belgian lung-cancer screening trial (NELSON). This was applied in eight types of cancer with the sensitivity
population-based, randomized, and controlled trial recruited ranging from 69% to 98% at a nearly 100% specificity .
[41]
30,959 persons with high-risk lung cancer in four selected CAPP-Se, which is the first NGS-based method for ctDNA
regions in the Netherlands and Belgium. The experimental quantitation, was invented. This technique searches for
group reduced 0.80 deaths per 1000 person-years after four exons harboring recurrent mutations in potential driver
rounds of LDCT screening compared with the control group genes from the catalogue of somatic mutations in cancer
at 10 years of follow-up, further revealing LDCT’s benefits in and is modified according to whole-exome sequencing of
high-risk people with lung cancer . These two influential NSCLC patients, to design the NCSLC selector. CAPP-
[38]
clinical trials consolidated LDCT as a reliable tool in early Seq has a sensitivity of 100% in Stage II-IV NSCLC with a
screening of lung cancer. specificity of 96% and was able to detect ctDNA in half of
[42]
However, health-care personnel face numerous clinical Stage I patients . The technique was further refined in the
problems using LDCT. First, LDCT has high false positive aspects of mutation spots selection, and a machine-learning
rate. The false positive rate of LDCT in NLST reached 96.4%, method termed “lung cancer likelihood in plasma” (Lung-
while the positive predictive value in NELSON test was only CLiP), which can robustly distinguish early stage NSCLC
43.5% [37,38] . Second, over-diagnosis causes unnecessary invasive from inflammation or benign nodules, was designed. In
treatments. A meta-analysis revealed that 1.7% negative patients with Stage I disease, Lung-CLip has a sensitivity of
[43]
positive patients received invasive diagnostic measures and 63% and a specificity of 80% .
serious complications happened in 0.4% of them . Third, Detection methods based on ctDNA’s fragmentomics
[39]
radiation exposure in the regular screening routine may act as characters were also examined in NSCLC screening.
the induction of other cancers. The effective dose from 256- Researchers developed an approach called DELFI by cfDNA
slice LDCT is 0.71 mSv each time, which is much lower than genome-wide analysis. This method was based on cancer-
that from regular CT examination. However, this extent of featured fragmentomics and achieved an AUC of 0.87 in lung
exposure still increases nominal lifetime intrinsic risks related cancer after machine learning data analysis . When further
[24]
to radiogenic cancer by 0.13% and 0.30% in male and female combined with clinical risk factors, carcinoembryonic
populations, respectively . Fourth, the targeted screening antigen levels, and CT imaging, DELFI identified 91%
[40]
population of the previous LDCT screening is restrained Stage I/II and 96% Stage III/IV lung cancer patients at the
to high-risk group, mostly smoking and aging population, specificity of 80% . ctDNA’s fragmentomics not only help
[44]
but the incidence of NSCLC in people free of these risks, are to screen lung cancer among healthy people, but also reveal
rising . For these low-risk population, whether LDCT brings the tissue origin of ctDNA. Fragment patterns, such as size
[1]
more benefits or harms remains unknown. distribution, fragment patterns, location of nucleosomes,
To improve the early screening coverage and accuracy, and open chromatin regions, have been proven as valid
alternative parameters from basic clinical information, characters to ascertain the tissue source of ctDNA [24,45-47] .
radiographic imaging, and traditional tumor markers Aberrant ctDNA methylation that precedes genomic
should be in place. Among them, ctDNA has gained wide mutation is an ideal biomarker for individuals with low
attention due to its non-invasiveness and high sensitivity, tumor burden, such as patients with early stage lung
[16]
which basically meet the need of early screening . cancer . A large-scale research in CCGA database
[16]
found whole-genome bisulfite sequencing outperformed
2.2. Application value of ctDNA in early detection CNVs and SNVs in identifying patients with cancers. The
Considering the genomic mutation in ctDNA, frequent detection pattern of targeted methylation meets the goal for
driver mutations in genes such as EGFR, KRAS, PIK3CA, population-level screening of more than 50 types of cancers
and TP53 and other less frequent mutations have been and is able to locate the origin at high accuracy, suggesting
investigated in the Ion PGM and AmpliSeq Cancer Panel . that ctDNA’s methylation is feasible in early detection .
[48]
[12]
About 60.3% of early-stage NSCLC patients were found to Chen et al. utilized nanoparticle-based DNA extraction
have detectable ctDNA and its concordance with matched (MOB) followed by qMSP to detect promotor methylation
tumor DNA was 50.4%. ctDNA further exhibited a higher on eight candidate lung cancer-specific genes . In a
[49]
positive prediction value than traditional protein biomarkers Chinese cohort, the three best methylation combination
in this study, which corroborated its potential as a promising model, containing CDO1, SOX17, and HOXA7, attained the

Volume 1 Issue 1 (2022) 6 https://doi.org/10.36922/gtm.v1i1.96