Biomedicine & Pharmacotherapy 67 (2013) 723–730

Available online at

www.sciencedirect.com

Original article

Circulating cell-free DNA, SLC5A8 and SLC26A4 hypermethylation,

BRAFV600E: A non-invasive tool panel for early detection of thyroid
cancer
Mariangela Zane a,1, Marco Agostini b,c,d,*,1, Maria Vittoria Enzo d, Eric Casal Ide a,
Paola Del Bianco e, Francesca Torresan a, Isabella Merante Boschin a, Gianmaria Pennelli f,
Andrea Saccani g, Domenico Rubello h, Donato Nitti d, Maria Rosa Pelizzo a
a
Surgical clinic II, Department of Surgical, Oncological and Gastroenterological Sciences, University of Padova, Padova, Italy
b
Department of Nanomedicine, The Methodist Hospital Research Institute, Houston, Texas, USA
c
Istituto di Ricerca Pediatrica-Città della Speranza, Padova, Italy
d
Surgical Clinic I, Department of Surgical, Oncological and Gastroenterological Sciences, University of Padova, via Nicolò Giustiniani, 2, 35128 Padova, Italy
e
Istituto Oncologico Veneto (IOV-IRCCS), Padova, Italy
f
Unit of Surgical Pathology and Cytopathology, Department of Medicine, University of Padova, Padova, Italy
g
EuroClone S.p.A, Pero (MI), Italy
h
Service of Nuclear Medicine, Department of Nuclear Medicine, PET/CT Centre, Radiology, Neuroradiology, Medical Physics, Santa Maria della Misericordia
Hospital, Rovigo, Italy

A R T I C L E I N F O A B S T R A C T

Article history: Purpose: In the latest years, high levels of circulating cell-free DNA (cf-DNA) have been found to be
Received 29 May 2013 associated with cancer diagnosis and progression, and cf-DNA has become a potential candidate as
Accepted 24 June 2013 biomarker for tumor detection. cf-DNA has been investigated in plasma or serum of many tumor patients
affected by different malignancies, but not yet in thyroid cancer (TC). Furthermore, in TC cells the
Keywords: capability to metabolize iodine is frequently lost. SLC5A8 and SLC26A4 genes are both involved in the
Cell-free-DNA iodine metabolism, and SLC5A8 hypermethylation status is associated with the BRAFV600E mutation,
Thyroid cancers
which is the most frequent genetic event underlying the development of papillary TC. The aim of our
Methylation
BRAF
study is the development of a new non-invasive tool for the diagnosis and prognosis of TC based on cf-
Early diagnosis DNA, SLC5A8 and SLC26A4 hypermethylation, and BRAFV600E analysis.
Methods: cf-DNA was measured by quantitative real-time PCR in nine cases of anaplastic thyroid cancer
(ATC), 58 medullary thyroid cancers (MTC), five of synchronous medullary and follicular thyroid cancers
(SMFC), 23 follicular adenomas (FA), 86 papillary thyroid cancers (PTC). A control group of 19 healthy
subjects was taken. Moreover, in the PTC group we analyze the state of hypermethylation of SLC5A8 and
SLC26A4, BRAFV600E mutation, and their involvement in the loss of function of the thyroid.
Results: cf-DNA showed a high ability to discriminate healthy individuals from cancer patients. cf-
DNAALU83 and cf-DNAALU244 values were significantly correlated with the histological type of TC (P-
value < 0.0001). A significant increase in the amount of cf-DNAALU83 and cf-DNAALU244 when
methylation occurs was observed (P-value = 0.02). A correlation between BRAFV600E and cf-
DNAALU244/ALU83 was also found (P-value = 0.02).
Conclusions: According to our experimental results, the panel including cf-DNA, SLC5A8 and SLC26A4
hypermethylation, and BRAFV600E analysis appears easy, reproducible, and non-invasive for the diagnosis
on TC. Its possible implication in clinical setting remains to be elucidated.
ß 2013 Elsevier Masson SAS. All rights reserved.

1. Introduction

Thyroid cancer (TC) is one of the few diseases that show a

constant increase in incidence, in particular in young adults (20–
* Corresponding author. Tel.: +39 049 821 2086/4374; fax: +39 049 651891.
44 years), 1.7 patients per 100,000 per year, while the annual
E-mail address: m.agostini@unipd.it (M. Agostini).
1
Mariangela Zane and Marco Agostini equally contributed to this work as first mortality rate is 0.7/100,000 and 0.3/100,000 per year respectively
author. for women and men (http://globocan.iarc.fr). TC differs worldwide

0753-3322/$ – see front matter ß 2013 Elsevier Masson SAS. All rights reserved.
http://dx.doi.org/10.1016/j.biopha.2013.06.007
724 M. Zane et al. / Biomedicine & Pharmacotherapy 67 (2013) 723–730
considerably: some of the highest incidence values have been
reported in Italy [1]. In young subjects (0–49 years), TC is the
second-most common tumor in women and the fifth in men –
excluding skin epitheliomas – occurring in Italy (http://www.re-
gistri-tumori.it/cms/). Currently, TC diagnosis is based only on
cytological clinical criteria, which are difficult to identify and made
at the onset of symptoms or during surgery. Despite the low
invasiveness, fine needle aspiration cytology (FNAC) is not a
method free of risk or complications. Moreover, FNAC needs
qualified medical staff and is performed in a protected area, and
this implies a performance more constrained than a simple blood
test. Research has helped to understand the factors involved in
thyroid cells malignant transformation; however, by now the large
quantity of proposed biomarkers has not significantly influenced
the clinical practice [2].
In recent years, high levels of circulating cell-free DNA (cf-DNA)
have been found to be associated with cancer diagnosis and
progression and have shown characteristics of potential candidate
as biomarker of cancer response [3–5]. cf-DNA has been
investigated in plasma or serum of many cancer patients affected
by different malignancies such as breast, prostate, bladder, ovarian
and colorectal cancers, but not yet in thyroid cancers [6].
Circulating cf-DNA is present in healthy subjects as a
physiologic event, but its increase has been identified as a strong
diagnostic and prognostic marker of malignancy [7–9]. Particular
attention should be spent to the proportions of cancer- and non-
cancer-derived cf-DNA, because confounding results have been
described [10]. Sources of cf-DNA were identified in apoptosis and
necrosis of cancer cells, that determine a release of nucleic acids
into the bloodstream. It has also been hypothesized that cf-DNA
released into circulation from cancer necrosis varied in size,
whereas cf-DNA released from apoptotic death is uniformly
truncated into fragments shorter than 200 bp [11–13]. In healthy
subjects, the main source of cf-DNA is thought to derive from
apoptotic cells; in contrast, necrotic cell death is a frequent event
in solid cancers and DNA fragments released from cancer cells are
variable in length [14–19].
cf-DNA is quantified by qPCR using the ALU repeats (i.e. the
most abundant repeat sequences in the human genome) in order to
evaluate its potential role as biomarker to discriminate cancer
patients from healthy subjects.
Several authors also revealed the presence of hypermethylation
gene events in circulating cf-DNA in different malignancies [20,21].
Gene methylation occurs on CpG nucleotides, concentrated in small
regions of DNA – CpG islands – grouped around the gene regulatory
region and can affect the transcriptional regulation of genes
themselves [22,23]. In some cancers, methylation can occur in
particular oncosuppressor genes or in genes involved in specific
organ function; in TC, the expression of molecules metabolizing
iodine is lost frequently [24–27]. Among the family of sodium-solute
symporters (SLC), SLC5A8 and SLC26A4 (PDS) encode for AIT and
Pendrin. They are genes for putative thyroid follicular cell apical
iodide transporters, that act in passive transport and diffusion and
are frequently methylated in papillary TC (PTC) (http://
www.ncbi.nlm.nih.gov/gene/160728; http://www.ncbi.nlm.nih.-
gov/gene/5172). Tumor suppressor function of SLC5A8, through
proapoptosis, has been documented; its methylation status has been
observed in different human cancers [28,29]. In TC, the hypermethy-
lation status of SLC5A8 has been associated with advanced tumors,
plurifocality, and extrathyroidal invasion [30]. Pendrin is localized at
the apical membrane of thyroid follicular cells and plays a crucial role
in thyroid metabolism by transporting iodine into the follicular
lumen, where accumulation, oxidation, and organification into
thyroglobulin occur [31,32]. Suppression of these thyroid iodine-
metabolizing molecules, through aberrant gene methylation, results
in the loss of the ability of cancer cells to concentrate iodine [33].
Methylation of SLC5A8 is associated with the BRAFV600E
mutation, which is the most frequent gene event at the basis of
the development of PTC [34–37]. In recent years, BRAFV600E
mutation has been used as an indicator of the presence of PTC
and has been observed in a percentage between 40 to 70% of the
examined cases [38,39]. This mutation leads to a more aggressive
behavior, presenting in an advanced stage at diagnosis and could
evolve towards a less differentiated histology [40], causing
genomic instability and an increased susceptibility of the mutated
cells to acquire additional defects [41].
In the present pilot study, we evaluated the presence of ALU
fragments in a large cohort of different histological types of TC
derived from follicular (papillary, follicular, and anaplastic
histotypes) and parafollicular cells (medullary form). Moreover,
in PTC patients we analyze the state of hypermethylation of SLC5A8
and SLC26A4 and its involvement in the loss of function of the
thyroid. Our goal was to provide a new non-invasive tool panel for
early detection of TC through the analysis of circulating cf-DNA, its
hypermethylation status and the presence of pathological muta-
tion in BRAF.
2. Materials and methods
2.1. Patients and tumor characteristics
The following study was conducted on a consecutive series of
181 patients with thyroid disease, stored in the Clinical Surgery I
Tissue Bank during the period between September 2007 and May
2011. We recruited patients who underwent total or near total
thyroidectomies in Unit of Clinical Surgery II, Department of
Surgical, Oncological and Gastroenterological Sciences (DiSCOG),
University of Padua, from whom had been collected plasma
samples. In details, we analyzed nine cases of anaplastic thyroid
cancer (ATC), 58 cases of medullary thyroid carcinoma (MTC), five
cases of synchronous medullary and follicular carcinoma (SMFC),
23 cases of follicular adenoma (FA) and 86 cases of papillary
thyroid carcinoma (PTC) of the classical form, including 42 at early
stage and 44 at late stage. From the medical records of each patient,
clinical and biological information were obtained, as medical
history, preoperative investigations, histological reports, follow-up
data, blood chemistry values, and molecular markers commonly
used in preclinical diagnosis, in particular: carcinoembryonic
antigen (CEA), anti-thyroglobulin antibodies (anti-TgAb) and anti-
thyroperoxidase antibodies (anti-TPOAb) to identify the presence
of thyroiditis, calcitonin (CT), and RET mutations (Table 1). In
parallel, we collected plasma samples by 19 healthy subjects which
were taken as controls.
The study protocol was reviewed and approved by the local
Ethics Committee (protocol number 448 P) and each patient
provided written informed consent.
2.2. Analysis of cf-DNA
2.2.1. Extraction of DNA from plasma and tumor tissue samples
Preoperation plasma samples were obtained on the day of
surgery. Ten millilitres of peripheral blood were drawn into a
purple-top blood collection tube (containing EDTA additive),
before physical examination or biopsy. Plasma samples were
obtained by centrifugation of the blood at 3000 g for 15 minutes,
within 6 hours of collection, then carefully collected from the
upper portion of the supernatant and stored in aliquots at 80 8C.
Biopsies of cancer tissue, obtained at the time of surgery, were
immediately frozen in nitrogen liquid and then stored at 80 8C.
cf-DNA was purified from 700 mL of plasma using a QIAmp1
UltraSensTM Virus Kit (Qiagen, Hilden, Germany), according to the
 M. Zane et al. / Biomedicine & Pharmacotherapy 67 (2013) 723–730 725

Table 1
Patients characteristics (number of cases analyzed).

FA PTC MTC SMFC ATC

n % n % n % n % n %

23 86 58 5 9

Medical history
Age (yr) Tot
Mean 51 45 49 61 70
Range (45–86) (56–92)
(31–73) (11–80) (5–83)
M
Mean 54 50 49 62 62
Range (45–86) (58–68)
(31–73) (11–68) (5–83)
F
Mean 50 43 49 61 74
Range (56–92)
(32–71) (12–80) (20–73) (60–62)

Sex M 9 39 26 30 29 50 3 60 3 33
F 14 61 60 70 29 50 2 40 6 67

Radiochemotherapy Yes 1 4 3 3 3 5 0 0 1 11
before the surgery No 22 96 81 94 53 91 5 100 7 78
NA 0 0 2 2 2 3 0 0 1 11

Thyroid function Eut 7 30 44 51 25 43 3 60 3 33

Ipot 3 13 12 14 3 5 0 0 2 22
Ipert 2 9 1 1 0 0 0 0 0 0
NA 11 48 29 34 30 52 2 40 4 44

Thyroiditis Yes 2 9 29 34 6 10 5 100 2 22

No 19 82 54 63 49 85 0 0 6 67
NA 2 9 3 3 3 5 0 0 1 11

Blood chemistry values

CT Pos 1 4 2 2 54 93 5 100 0 0
Neg 12 52 26 30 4 7 0 0 3 33
NA 10 43 58 67 0 0 0 0 6 67

CEA Pos 0 0 0 0 21 36 3 60 0 0
Neg 2 9 2 2 5 9 2 40 0 0
NA 21 91 84 98 32 55 0 0 9 100

Anti-TGAb Pos 2 9 22 26 3 5 0 0 1 11
Neg 6 26 35 41 23 40 4 80 3 33
NA 15 65 29 34 32 55 1 20 5 56

Anti-TPOAb Pos 0 0 20 23 6 10 0 0 1 11
Neg 7 30 36 42 21 36 3 60 3 33
NA 16 70 30 35 31 54 2 40 5 56

RET mutation Mut 0 0 0 0 37 64 0 0 0 0

wt 2 9 0 0 19 33 5 100 0 0
NA 21 91 86 100 2 3 0 0 9 100

Follow-up
131
I therapy Yes 1 4 74 86 0 0 2 40 3 33
No 0 0 7 8 0 0 0 0 1 11
NA 22 96 5 6 58 100 3 60 5 56

TG post Pos 1 4 27 31 0 0 0 0 2 22
Neg 0 0 54 63 0 0 2 40 1 11
NA 22 91 5 6 58 100 3 60 6 67

Anti-TGAb post Pos 0 0 10 12 0 0 0 0 1 11

Neg 1 4 71 82 0 0 2 40 2 22
NA 22 96 5 6 58 100 3 60 6 67

Radiochemotherapy Yes 0 0 1 1 0 0 0 0 3 33
No 0 0 0 0 0 0 0 0 0 0
NA 0 0 85 99 58 100 5 100 6 67

PET Yes 0 0 0 0 0 0 0 0 2 22
No 0 0 0 0 0 0 0 0 1 11
NA 0 0 86 100 58 100 5 100 6 67

CT post Pos 1 4 0 0 33 57 3 60 0 0
Neg 1 4 0 0 23 40 2 40 0 0
NA 21 91 0 0 2 3 0 0 9 100

CEA post Pos 0 0 0 0 9 16 1 20 0 0

Neg 1 4 0 0 16 28 1 20 0 0
NA 22 96 0 0 33 57 3 60 9 100
726 M. Zane et al. / Biomedicine & Pharmacotherapy 67 (2013) 723–730
Table 1 (Continued )
FA PTC
n % n %
23 86
OS n 0 0 80 93
Mean (mo.) – 28,6
Range (mo.) – [2;47]
Disease-free survival AWD 0 0 7 8
DOD 0 0 2 2
NED 0 0 71 83
NA 23 100 6 7
FA: follicular adenoma; PTC: papillary thyroid carcinoma; MTC: medullary thyroid cancer; SMFC: synchronous medullary and follicular carcinoma; ATC: anaplastic thyroid
cancer; CT: calcitonin; CEA: carcinoembryonic antigen; anti-TgAb: anti-thyroglobulin antibody; anti-TPOAb: anti-thyroperoxidase antibody; OS: overall survival.
manufacturer’s instructions. cf-DNA preparations were eluted in
40 mL of buffer elution. Tumor biopsies were homogenized by
high-speed agitation with beads, then DNA was isolated using a
QIAamp1 DNA Mini Kit (Qiagen, Hilden, Germany). Eluted cf-DNA
and tissue DNA were stored at 20 8C, until further use.
2.2.2. Quantitative PCR of plasma DNA fragments
The quantification of the cf-DNA fragments was performed by
quantitative real-time PCR (qPCR), which amplified and quantified
the shortest and longest DNA fragments. To maximize the
sensitivity of cf-DNA quantification, the ALU repeats (which are
the most abundant repeat sequences in the human genome) were
used as a target of the qPCR.
Two primer pairs were designed as follows: in one set of
primers (ALU83) were amplified both the shortest (83 bp) and
longest fragments, whereas in the second set of primers (ALU244)
were amplified only the longest (244 bp) DNA fragments. The
results obtained using the ALU83 primers represent the total free
plasma DNA, while the results obtained using the ALU244 primers
reflect the amount of DNA released from non-apoptotic cells. The
quantitative values from the 83 bp amplicon represent the total
level of cf-DNA in ng/mL, while the ratio of longer to shorter
fragments demonstrates the integrity of cf-DNA in each sample.
The sequences of the ALU83 primers were as follows: forward,
5’-CTGAGGTCAGGAGTTCGAGACC-3’, and reverse, 5’-CCACGCC-
CGGCTAATTTT-3’; the ALU244 primers were as follows: forward,
5’-GCGGTGGCTCACGCCTGTAA-3’, and reverse, 5’-GGAGTGCAGT-
GGCGCGATCT-3’. The probe sequence was 6FAM-CCTGGCCAA-
CATGGTGAAACCCC-TMR.
The reaction was performed using the 7300 Real-Time PCR System
and 7500FastReal-TimePCRSystem(AppliedBiosystems,Milan,Italy).
The reaction was performed in 20 mL reaction volumes containing 2X
FluoCycleTM II,probe(Euroclone1,Milan,Italy),0.9 mMofeachALU244
primerand0.05 mMofforwardand0.9 mMofreverseprimerforALU83,
0.25 mM of probe and 1 mL of cf-DNA. The real-time PCR conditions
consisted of an initial denaturation step for 5 minutes at 95 8C followed
by 40 cycles of denaturation for 15 seconds at 95 8C and annealing/
extension for 1 minute at 62 8C. The appropriate internal and negative
referencesampleswereincluded.Theabsoluteamountofcf-DNAineach
sample was determined by a standard curve, using 10-fold dilutions
(from10 ngto100 fg)ofgenomicDNAfromperipherallymphocytesofa
healthy subject.
The quantitative ratio of ALU244 to ALU83 reflected the
integrity of the cf-DNA (integrity index = cf-DNAALU244/ALU83).
2.3. Analysis of cf-DNA methylation
2.3.1. Sodium bisulfite treatment of DNA
Plasma cf-DNA was modified with a sodium bisulfite method,
using an EpiTect1 Bisulfite Kit (Qiagen, Hilden, Germany)
following the manufacturer’s protocol. For bisulfite conversion,
MTC SMFC ATC
n % n % n %
58 5 9
56 100 5 100 9 100
21,4 20,8 16,2
[0;51] [8;34] [3;49]
20 35 1 20 7 78
3 5 1 20 2 22
33 57 3 60 0 0
2 3 0 0 0 0
we added Bisulfite Mix Solutions and DNA Protect Buffer to 40 mL
of cf-DNA in a 0.2 ml tube and performed the conversion using
VeritiTM Thermal Cycler (Applied Biosystems, Milan, Italy) for 5 h
at 60 8C with three thermal spikes at 95 8C. We purified the
converted single stranded DNA, using an Epitect Spin Column
based purification kit for bisulfite-converted DNA (Qiagen, Hilden,
Germany). Purified modified cf-DNA was eluted in 25 mL of Elution
Buffer and PCR analysis directly used in real-time.
2.3.2. Quantitative real-time PCR methylation
Sodium bisulfite-treated plasma cf-DNA was analyzed by a
primer-probe set using a 7300 Real-Time PCR System and a 7500
Fast Real-Time PCR System (Applied Biosystems, Milan, Italy).
Briefly, three sets of primers and probes designed specifically
for bisulfite-converted DNA were used: two sets for the genes of
interest (SLC5A8, SLC26A4) and one set for MYOD, which was used
to control the bisulfite conversion and for normalization of the
amount of input cf-DNA. Primers and probes for SLC5A8, SLC26A4
and MYOD were as follows; the SLC5A8 forward primer was 5’-
TTGCGCGTTAGGGATTCG-3’, the SLC5A8 reverse primer was 5’-
CCCGTAACGTATCCATAACCG-3’, and the SLC5A8 probe was 6FAM-
GATCGCCTAAACCCTACG-TMR; the SLC26A4 forward primer was
5’-AGCGTTAGGCGGTAGGTCG-3’, the SLC26A4 reverse primer was
5’-TCTCCCGCAACGTCTTACG-3’, and the SLC26A4 probe was 6FAM-
CGGTCGGTTTATAGCGAGT-TMR; the MYOD forward primer was 5’-
TTTTAGTTAGAGTGTTGAGAGGATTGTGT-3’, the MYOD reverse pri-
mer was 5’-CATACCGACCACCCCCATAA-3’, and the MYOD probe
was 6FAM-5’-ATTGTAGATTTAGGAAGAGGTT-3’-TMR.
The reaction was performed in 20 mL reaction volumes,
containing 1X TaqMan Universal PCR Master Mix No AmpErase
UNG, 0.9 mM of each SLC26A4 and MYOD primer and 0.3 mM of
forward and 0.05 mM of reverse primer for SLC5A8, 0.25 mM of
probe for each gene and 4 mL of modified cf-DNA. The real-time
PCR conditions consisted of an initial denaturation step for
10 minutes at 95 8C followed by 50 cycles of denaturation for
15 seconds at 95 8C and annealing/extension for 1 minute at 60 8C.
The appropriate negative reference samples were included:
without template, non-methylated control (Qiagen, Germany)
and non-modified template.
Absolute quantity was derived from a standard curve generated
by serial dilution of fully Methylated Control DNA (Qiagen,
Germany), which also confirmed the specificity of the reactions
for methylated DNA. The concentration of plasma cf-DNA was
expressed as ng/ml.
The amount of methylated cf-DNA at a specific locus, called
Percentage of Methylated Reference (PMR), was calculated by
dividing the GENE:MYOD ratio of a sample by the GENE:MYOD ratio
of the fully methylated sample and multiplying by 100 [36].
Reactions using fully Methylated Control DNA were applied to
normalize any differences in amplification efficiency between the
target genes and MYOD.
 M. Zane et al. / Biomedicine & Pharmacotherapy 67 (2013) 723–730 727

A gene was deemed methylated if the percentage of PMR value
was greater than 0 in the case of cf-DNA and greater than 1 for
tissue DNA. cf-DNA samples were regarded as methylated if their
PMR value was greater than 0 at least in one of two genes. For the
tissue samples, methylated samples presenting a PMR value
greater than or equal to 1% were considered.
area under the curve (AUC), the 95% confidence intervals (95% CI)
and the level of significance (P-value < 0.05).
3. Results
3.1. Patient, tumor and histopathology characteristics

2.4. Analysis of BRAFV600E mutation
Plasma cf-DNA and tumor tissue DNA were analyzed with the
High Resolution Melting Assay (HRMA) (PREMIER Biosoft Inter-
national, Palo Alto CA, USA) to investigate the mutation V600E in
the BRAF gene. The tumor tissue sample was available for 68 cases
of PTC.
Primer sequences for exon 15 of BRAF in HRMA were 5’-
ATGAAGACCTCACAGTAAAAATAG-3’ (forward) and 5’-GACAACT-
GTTCAAACTGATGGTGG-3’ (reverse) for the shorter 88 bp
amplicon.
Thirty ng of tumor DNA and 2 mL of cf-DNA were amplified in a
final volume of 20 mL by using the following protocol: 1X
AmpliTaq Gold PCR Master Mix (Applied Biosystems, Milan)
which includes 2.5 mM MgCl2 and 0.2 mM of each dNTPs, 1.5 mM
of SYTO 9 (Invitrogen, Oregon, USA) and 0.3 mM of each primer.
All PCR reactions were performed in single, including a wild-
type and a mutation DNA reference samples previously sequenced.
PCR cycling and HRM analysis were performed on the 7500 Fast
Real-Time PCR System (Applied Biosystems, Milan). The amplifica-
tion was performed in an initial denaturation at 95 8C for
10 minutes followed by 40 cycles of 95 8C for 15 seconds, 58 8C
for 15 seconds, 60 8C for 15 seconds and a melt from 60 to 95 8C
rising at 0.3 8C per second.
The analysis of dissociation curves was performed using a HRM
v2.0 Software.
2.5. Statistical analysis
The results were subjected to statistical analysis using the
Kruskal-Wallis test. To investigate the effectiveness of the markers
in the disease detection, a comparison was made between two
main groups: the first consisting of 181 investigated patients with
thyroid disease, and the second consisting of 19 healthy subjects.
For the detection of total circulating cf-DNA, we investigated the
amount of DNA amplified with the assay designed for the ALU83:
the distributions of circulating DNA, found in patients and healthy
subjects, were compared using the Kruskal-Wallis test.
The capability to discriminate healthy individuals from thyroid
cancer patients of the different methods of cf-DNA quantification
was investigated also by the ROC curve analysis, calculating the
Among 181 patients included in the study group, thyroid
disease histotype was as follows: n = 23 FA, n = 9 ATC, n = 58 MTC,
n = 5 SMFC, and n = 86 PTC. The cancer patients were clinically
staged according to the seventh edition of TNM system for TC
designed by the American Joint Committee on Cancer (AJCC). The
patients underwent surgery; specifically, they received total
thyroidectomy (n = 157, 87%), lobectomy (n = 13, 1%), and/or
lymph node dissection (n = 164, 91%). Details of the patient,
tumor, histopathology and staging are summarized in Tables 1
and 2.
3.2. Plasma cf-DNA distribution
3.2.1. Plasma cf-DNA distribution in healthy subjects and TC patients
The ability of cf-DNA to discriminate among TC patients and
healthy subjects was confirmed using a control group of plasma
samples from healthy donors matched for age and gender; the
samples were obtained immediately after careful exclusion of
disease presence. Among patients’ groups, the baseline levels of cf-
DNAALU83 and cf-DNAALU244 were significantly higher than in the
control group (cf-DNAALU83 median: patients, 22.54 ng/mL; range
1.48–3760.00 ng/mL; controls, 5.14 ng/mL; range 1.07–13.13 ng/
mL; P-value < 0.0001; – cf-DNAALU244; patients: median 6.36 ng/
mL, range 0.31–758.83 ng/mL; controls: median 2.64; range 0.30–
4.41 ng/mL; P-value < 0.0001).
By ROC curve analysis, each method of cf-DNA quantification
showed a high ability to discriminate healthy subjects from TC
patients (AUCALU83: 0.91, 95% CI: 0.86–0.94; AUCALU244: 0.84, 95%
CI: 0.78–0.89, Fig. 1). Adoptying a cut-off of 11.741 ng/mL, the
diagnostic power of cf-DNAALU83 showed 94.7% specificity (95% CI:
74.0–99.9) and 73.5% sensitivity (95% CI: 66.4–79.8); with a cut-off
of 4.412 ng/mL, the diagnostic power of cf-DNAALU244 showed
100% specificity (95% CI: 82.4–100) and 67% sensitivity (95% CI:
59.5–73.7).
3.2.2. Plasma cf-DNA and pathology characteristics of thyroid cancers
Median plasma cf-DNAALU83 and cf-DNAALU244 were signifi-
cantly correlated with the histological type of TC. The less
aggressive follicular derivation tumors, FA and PTC, showed a total
cf-DNAALU83 median higher than more advanced undifferentiated
cancers ATC, and than parafollicular derivation carcinomas

Table 2
Tumor, histopathology and staging characteristics.

TNM

Group T N M Tumor stage

1 2 3 4a 4b nd 0 1 1a 1b nd 0 1 0 1 2 3 4A 4B 4C

FA n 23 – – – – – – – – – – – – – – – – – – – –
% – – – – – – – – – – – – – – – – – – – –

PTC n 86 16 9 58 3 0 – 37 – 30 19 85 1 – 42 0 34 10 0 0
% 100 19 10 67 4 0 – 43 – 35 22 99 1 – 49 0 40 12 0 0

MTC n 58 24 2 14 2 0 16 32 1 2 11 12 58 0 3 24 2 7 10 0 0
% 100 41 3 24 3 0 28 55 2 3 19 21 100 0 8 41 3 12 17 0 0

SMFC n 5 3 1 1 0 0 – 1 3 1 – 5 0 – 1 0 3 1 0 0
% 100 60 20 20 0 0 – 20 60 20 – 100 0 – 20 0 60 20 0 0

ATC n 9 0 1 1 3 3 1 4 2 1 1 1 8 1 – 0 0 1 3 3 1
% 100 0 11 11 33 33 11 44 22 11 11 11 88 11 – 0 0 11 33 33 11
728 M. Zane et al. / Biomedicine & Pharmacotherapy 67 (2013) 723–730

Fig. 1. ROC cruve for discriminating TC patients from healthy subjects using (a) cf-
DNAALU83 plasma level and (b) cf-DNAALU244 plasma level.

MTC. The same trend was observed for cf-DNAALU244. As expected,

SMFC has a trend that is intermediate between PTC and MTC,
being composed of both follicular and parafollicular cells (Fig. 2a,
b). Calculating the ratio between the amount of cf-DNAALU244 and
cf-DNAALU83, we observed that SMFC and MTC showed a median
value higher than FA and PTC, but the group with the highest value
was ATC (Fig. 2c).
There was a statistically significant correlation between cf-
DNAALU83 and tumor stage in PTC samples, with an increase of
median value from 20.67 ng/mL in the early stage (I) to 37.45 ng/
mL for late stages (III, IV) (early stage cf-DNAALU83 range 1.90–
548.57 ng/mL; late stage cf-DNAALU83 range 7.89–3760 ng/mL; P-
value 0.0094), while there was no correlation between cf-
Fig. 2. Box-plot of (a) cf-DNAALU83 vs TC histotype, (b) cf-DNAALU244 vs TC histotype
DNAALU244 and tumor stage in PTC cases (early stage cf-DNAALU244 and (c) cf-DNAALU244/ALU83 vs TC histotype (FA: follicular adenoma; PTC: papillary
range: 0.33–229.26 ng/mL; late stage cf-DNAALU244 range: 1.85– thyroid cancer; MTC: medullary thyroid carcinoma; SMFC: synchronous medullary
58.83 ng/mL; P-value 0.0710). Similar, even if not statistically and follicular carcinoma; ATC: anaplastic thyroid cancer).
significant, is the relationship between MTC samples and
 M. Zane et al. / Biomedicine & Pharmacotherapy 67 (2013) 723–730
cf-DNAALU83, with a median value of 8.90 ng/mL for early stage I–II
(range: 1.48–1954.29 ng/mL) and 17.68 ng/mL for late stage III–IV
(range: 5.88–44.99 ng/mL, P-value 0.0590). No relationship
between cf-DNAALU244 and tumor stage in MTC was observed.
The number of ATC and SMFC groups (ATC n = 9, SMFC n = 5) were
considered too small to permit an adequate statistical analysis. No
relationship among the primary tumor and metastasis and cf-DNA
amount was observed in all histotype groups.
3.2.3. cf-DNA plasma, molecular markers and clinical features of
thyroid cancers
For the analysis of molecular markers, we observed a
correlation between Ct positive value and cf-DNA; positive values
of Ct correspond to a lower amount of cf-DNA (cf-DNAALU83: CTpos
11.70 ng/mL; CTneg 30.34 ng/mL; P-value < 0.0001 – cf-DNAALU244:
CTpos 4.39 ng/mL; CTneg 8.26 ng/mL; P-value, 0.0002). Similar trend
was observed between RET mutation status and cf-DNAALU244
(RETmut 2.79 ng/mL; RETwt 5.06; P-value 0.0402), while no
statistical significance was found between cf-DNAALU83 and RET
mutation status. These analyses suggest the possibility of having,
with cf-DNA, a potential substitute marker of the MTC presence in
the case that the classic markers, such as CT and RET, are negative.
No correlation was observed between cf-DNAALU83, cf-DNAALU244
and CEA, anti-TGAb, and anti-TPOAb.
Moreover, no correlation has been observed between cf-DNA and
clinical features, such as plurifocality, microcarcinomas, thyroid
function, thyroiditis, iodine therapy before surgery, sex or age.
3.3. Analysis of methylation of SLC5A8 and SLC26A4 genes
3.3.1. cf-DNA methylation events in cf-DNA
Results of SLC5A8 and SLC26A4 methylation events showed very
low values in the amount of methylated DNA, probably due to the
reduced amount of cf-DNA in plasma and its high level of
fragmentation. The results reported the presence of gene
methylation in tissue samples in 35 out of 86 cases of PTC
(40%), in particular 34 PTC with SLC5A8 hypermethylation events
and one PTC with SLC26A4 hypermethylation event, and in plasma
samples in seven out of 80 cases of PTC (9%), in particular six PTC
with SLC5A8 hypermethylation events and one PTC with SLC26A4
hypermethylation event (Table 3).
We observed a considerable increase of the amount of cf-
DNAALU83 when methylation occurs, with statistical significance; the
methylated cf-DNA group median value was about nine times higher
than the non-methylated cf-DNA group (cf-DNAunmeth median:
33.23 ng/mL; cf-DNAmeth median: 322.91 ng/mL; P-value 0.0242).
The same trend was observed for cf-DNAALU244, with a median value
approximately 13 times higher in the group with methylation events
than those where methylation did not occur (cf-DNAunmeth median:
8.43 ng/mL; cf-DNAmeth median: 109.58; P-value 0.0231).
3.3.2. Tissue BRAFV600E mutation status and cf-DNA plasma
The identification of the mutation BRAFV600E was possible only
in tumor tissue DNA but not in cf-DNA because of the difficulty of
recovery of cf-DNA. The mutation BRAFV600E was found in 46 out of
Table 3
Analysis of methylation of SLC5A8 and SLC26A4 genes.
SLC5A8 SLC26A4 Both
cf-DNA Tumor cf-DNA Tumor cf-DNA
Methylated n 6 34 1 1 7 35
% 7 40 1 1 8 41
Unmethylated n 80 52 85 85 79
% 93 60 99 99 92
729
68 tumor tissue samples, corresponding to 68% of cases. No
correlation between cf-DNAALU83 and cf-DNAALU244 and BRAFV600E
was observed; a correlation between BRAF mutation and cf-
DNAALU244/ALU83 was found with statistical significance (BRAFwt
median: 0.24; BRAFV600E median: 0.33; P-value 0.02).
4. Discussion
Substantial problems of TC are detection time and the methods
to obtain an early diagnosis: currently, there are no non-invasive
tests that can identify the presence of tumor and give unequivocal
results [37]. Furthermore, the incidence of TC is increasingly
affecting young people, who need efficient and specific screening
methods for this disease [1]. Surgical and pharmacological
therapies can effectively cure early-stage cancers, while disease
identified in advanced stages is usually difficult to treat effectively.
Fluid sampling is the easiest way to investigate the presence of
malignancies; it is a low-cost non-invasive kind of test and widely
accepted [42]. For that reason, our intention is to elaborate a tool
panel through the analysis of circulating cf-DNA amount, its
hypermethylation status and the presence of pathological BRAF
mutation.
The release of cf-DNA into the bloodstream has been
hypothesized to be due to different mechanisms, such as apoptosis,
necrosis and secretion, reflecting both physiological and patholo-
gical processes [43]. Thus, during tumor progression and cell
turnover, cf-DNA can be found of both normal and tumor origin in
the blood. High cf-DNA yields have been found in patients with
inflammatory diseases too [20], but in our case studies, there was
no correlation between cf-DNA amount and thyroiditis. Using the
panel test for thyroid malignancies we proposed, there was the
possibility to discriminate cancer patients from healthy subjects
using a cut-off cf-DNA amount for both cf-DNAALU83 and cf-
DNAALU244 assays. Moreover, it was possible to distinguish the
presence of benign tumors from cancer in cases groups, i.e.,
adenomas from cancers. Yet, our results showed that late stages
PTC are characterized by a cf-DNAALU83 higher than in early stage
PTC, suggesting a worse in tumor biology and behaviour. We could
also discriminate TC cancer of different subtypes: follicular,
parafollicular, papillary and anaplastic and medullary cancers.
This is a novelty in the diagnostic field for thyroid pathology and
represents an aid to study and better understands the molecular
mechanisms underlying thyroid cancerogenesis. Analyzing the
ratio between cf-DNAALU244 and cf-DNAALU83, we should identify
the non-apoptotic derivation DNA amount on the total cf-DNA
quantity. Follicular cell differentiated tumors, i.e., FA and PTC,
show an amount of cf-DNAALU83 higher than parafollicular cell
cancers (MTC and SMFC), but the lowest ratio of cf-DNAALU244/
ALU83; this suggests that follicular cell derivation tumors are more
prone to the release of cf-DNA compared to parafollicular cell
derivation malignancies, and the manner of release is programmed
cell death. Among follicular cell derivation cancers, there is a
consistent difference considering cases with different histopatho-
logic grades; FA and PTC release a larger amount of cf-DNAALU83
compared to undifferentiated ATC, but the latter has a higher cf-
DNA portion of non-apoptotic derivation (cf-DNAALU244/ALU83). This
indicates a higher necrotic component in cancers with a low grade
of differentiation compared to less differentiated tumors. Defini-
tively, a high yield of cf-DNA seems to correspond to the presence
Tumor of more aggressive and less differentiated malignancies. Another
important point is the variability within each histotype. Healthy
subjects have a cf-DNAALU83 distribution within a very narrow
range (on the order of 101), the adenomas patients in a range a little
51
wider (on the order of 102), and PTC and MTC patients in a
59
markedly wider range (on the order of 103), while ATC and SMFC
730 M. Zane et al. / Biomedicine & Pharmacotherapy 67 (2013) 723–730
groups contain a number of cases that are too low to consider their
variability. This difference between the various groups allows
establishing a cut-off above which there is definitely the presence
of disease. Moreover, the results support the concept of genetic
variability within the same cancer patient class for cancers derived
both from follicular and parafollicular cells. Tumors are the result
of a mix of different cell clones, which lead to the genetic and
epigenetic heterogeneity of cancers. Many studies have reported
the presence of epigenetic events in plasma or serum of cancer
patients, in particular DNA methylation of gene promoter regions
[21]. In our study, we found a considerable increase of cf-DNA in
the presence of aberrant methylation of SLC5A8 and/or SLC26A4, for
both cf-DNAALU83 and cf-DNAALU244 assays. This indicates a strong
correlation between aberrant methylation and the release of
circulating cf-DNA in TC patients, suggesting a crucial role of loss of
function of iodide-metabolizing molecules in thyroid carcinogen-
esis. The mutation BRAFV600E is present in 68% of PTC cases
analyzed according to the literature [39]. We found a significant
correlation between BRAFV600E and cf-DNAALU244/ALU83, meaning
that patients with a greater ratio are most likely to encounter the
mutation in BRAF than the others. BRAFV600E was seen to have a role
in tumorigenesis and in the dedifferentiation of PTC, as well as a
possible correlation with increased aggressiveness of the tumor.
The combination of these two factors may affect the normal
mechanisms of apoptosis and trigger necrotic processes.
As has already been demonstrated in other cancers, the
quantification of cf-DNA has proved to play an important role
for TC as well. In addition, methylation analysis alongside cf-DNA
assay has made identification of tumors and their locations more
specific. Accordingly to our experimental preliminary results, the
panel including cf-DNA, SLC5A8 and SLC26A4 hypermethylation,
and BRAFV600E analysis appears easy, reproducible, and non-
invasive for the diagnosis on TC. Its possible implication in clinical
setting remains to be elucidated.
Disclosure of interest
The authors declare that they have no conflicts of interest
concerning this article.
Acknowledgment
This article is dedicated to Leo Izzi.
This study was supported in part by AIRC Foundation,
EuroClone S.p.A. and I.O.V. Biological samples were provided by
Clinical Surgery I Tissue Bank, Tumor Tissue Biobank, Unit of
Clinical Surgery II.
References
[1] Lise M, Franceschi S, Buzzoni C, et al. Changes in the incidence of thyroid cancer
between 1991 and 2005 in Italy: a geographical analysis. Thyroid 2012;22:27–34.
[2] Nikiforov YE. Molecular diagnostics of thyroid tumors. Arch Pathol Lab Med
2011;135:569–77.
[3] Agostini M, Pucciarelli S, Enzo MV, et al. Circulating cell-free DNA: a promising
marker of pathologic tumor response in rectal cancer patients receiving
preoperative chemoradiotherapy. Ann Surg Oncol 2011;18:2461–8.
[4] Anker P, Lyautey J, Lederrey C, Stroun M. Circulating nucleic acids in plasma or
serum. Clin Chim Acta 2001;313:143–6.
[5] De Mattos-Arruda L, Olmos D, Tabernero J. Prognostic and predictive roles for
circulating biomarkers in gastrointestinal cancer. Future Oncol 2011;7:1385–97.
[6] Alix-Panabieres C, Schwarzenbach H, Pantel K. Circulating tumor cells and
circulating tumor DNA. Annu Rev Med 2012;63:199–215.
[7] Taback B, Hoon DS. Circulating nucleic acids and proteomics of plasma/serum:
clinical utility. Ann N Y Acad Sci 2004;1022:1–8.
[8] Taback B, Hoon DS. Circulating nucleic acids in plasma and serum: past,
present and future. Curr Opin Mol Ther 2004;6:273–8.
[9] Fujimoto A, O’Day SJ, Taback B, Elashoff D, Hoon DS. Allelic imbalance on
12q22-23 in serum circulating DNA of melanoma patients predicts disease
outcome. Cancer Res 2004;64:4085–8.
[10] Jung K, Fleischhacker M, Rabien A. Cell-free DNA in the blood as a solid tumor
biomarker–a critical appraisal of the literature. Clin Chim Acta 2010;411:1611–24.
[11] Gormally E, Caboux E, Vineis P, Hainaut P. Circulating free DNA in plasma or
serum as biomarker of carcinogenesis: practical aspects and biological signifi-
cance. Mutat Res 2007;635:105–17.
[12] Jahr S, Hentze H, Englisch S, et al. DNA fragments in the blood plasma of cancer
patients: quantitations and evidence for their origin from apoptotic and
necrotic cells. Cancer Res 2001;61:1659–65.
[13] Giacona MB, Ruben GC, Iczkowski KA, Roos TB, Porter DM, Sorenson GD. Cell-
free DNA in human blood plasma: length measurements in patients with
pancreatic cancer and healthy controls. Pancreas 1998;17:89–97.
[14] Deligezer U, Eralp Y, Akisik E, et al. Size distribution of circulating cell-free DNA
in sera of breast cancer patients in the course of adjuvant chemotherapy. Clin
Chem Lab Med 2008;46:311–7.
[15] Taback B, O’Day SJ, Boasberg PD, et al. Circulating DNA microsatellites:
molecular determinants of response to biochemotherapy in patients with
metastatic melanoma. J Natl Cancer Inst 2004;96:152–6.
[16] Umetani N, Giuliano AE, Hiramatsu SH, et al. Prediction of breast tumor
progression by integrity of free circulating DNA in serum. J Clin Oncol
2006;24:4270–6.
[17] Jiang WW, Zahurak M, Goldenberg D, et al. Increased plasma DNA integrity
index in head and neck cancer patients. Int J Cancer 2006;19:2673–6.
[18] Pinzani P, Salvianti F, Zaccara S, et al. Circulating cell-free DNA in plasma of
melanoma patients: qualitative and quantitative considerations. Clin Chim
Acta 2011;412:2141–5.
[19] Umetani N, Kim J, Hiramatsu S, et al. Increased integrity of free circulating DNA
in sera of patients with colorectal or periampullary cancer: direct quantitative
PCR for ALU repeats. Clin Chem 2006;52:1062–9.
[20] Kohler C, Barekati Z, Radpour R, Zhong XY. Cell-free DNA in the circulation as a
potential cancer biomarker. Anticancer Res 2011;31:2623–8.
[21] Fleischhacker M, Schmidt B. Circulating nucleic acids (CNAs) and cancer–a
survey. Biochim Biophys Acta 2007;1775:181–232.
[22] Shames DS, Minna JD, Gazdar AF. Methods for detecting DNA methylation in
tumors: from bench to bedside. Cancer Lett 2007;251:187–98.
[23] XingM.Genemethylationinthyroidtumorigenesis.Endocrinology2007;148:948–53.
[24] Sheils OM, Sweeney EC. TSH receptor status of thyroid neoplasms–TaqMan
RT-PCR analysis of archival material. J Pathol 1999;188:87–92.
[25] Venkataraman GM, Yatin M, Marcinek R, Ain KB. Restoration of iodide uptake
in dedifferentiated thyroid carcinoma: relationship to human Na+/I-sympor-
ter gene methylation status. J Clin Endocrinol Metab 1999;84:2449–57.
[26] Ringel MD, Anderson J, Souza SL, et al. Expression of the sodium iodide
symporter and thyroglobulin genes are reduced in papillary thyroid cancer.
Mod Pathol 2001;14:289–96.
[27] Arturi F, Russo D, Bidart JM, Scarpelli D, Schlumberger M, Filetti S. Expression
pattern of the pendrin and sodium/iodide symporter genes in human thyroid
carcinoma cell lines and human thyroid tumors. Eur J Endocrinol 2001;145:129–35.
[28] Li H, Myeroff L, Smiraglia D, et al. SLC5A8, a sodium transporter, is a tumor
suppressor gene silenced by methylation in human colon aberrant crypt foci
and cancers. Proc Natl Acad Sci U S A 2003;100:8412–7.
[29] Ganapathy V, Gopal E, Miyauchi S, Prasad PD. Biological functions of SLC5A8, a
candidate tumour suppressor. Biochem Soc Trans 2005;33:237–40.
[30] Hu S, Liu D, Tufano RP, et al. Association of aberrant methylation of tumor
suppressor genes with tumor aggressiveness and BRAF mutation in papillary
thyroid cancer. Int J Cancer 2006;119:2322–9.
[31] Twyffels L, Massart C, Golstein PE, et al. Pendrin: the thyrocyte apical mem-
brane iodide transporter? Cell Physiol Biochem 2011;28:491–6.
[32] Xing M, Tokumaru Y, Wu G, Westra WB, Ladenson PW, Sidransky D. Hyper-
methylation of the Pendred syndrome gene SLC26A4 is an early event in
thyroid tumorigenesis. Cancer Res 2003;63:2312–5.
[33] Catalano MG, Fortunati N, Boccuzzi G. Epigenetics modifications and thera-
peutic prospects in human thyroid cancer. Front Endocrinol 2012;3:40.
[34] Porra V, Ferraro-Peyret C, Durand C, et al. Silencing of the tumor suppressor
gene SLC5A8 is associated with BRAF mutations in classical papillary thyroid
carcinomas. J Clin Endocrinol Metab 2005;90:3028–35.
[35] Karasarides M, Chiloeches A, Hayward R, et al. B-RAF is a therapeutic target in
melanoma. Oncogene 2004;23:6292–8.
[36] Kondo T, Ezzat S, Asa SL. Pathogenetic mechanisms in thyroid follicular-cell
neoplasia. Nat Rev Cancer 2006;6:292–306.
[37] Xing M. Recent advances in molecular biology of thyroid cancer and their
clinical implications. Otolaryngol Clin North Am 2008;41:1135–46.
[38] Pelizzo MR, Boschin IM, Barollo S, et al. BRAF analysis by fine needle aspiration
biopsy of thyroid nodules improves preoperative identification of papillary
thyroid carcinoma and represents a prognostic factor. A mono-institutional
experience. Clin Chem Lab Med 2010;49:325–9.
[39] Giusti F, Falchetti A, Franceschelli F, Marini F, Tanini A, Brandi ML. Thyroid
cancer: current molecular perspectives. J Oncol 2010;17. http://dx.doi.org/
10.1155/2010/351679 [Article ID 351679].
[40] Durante C, Puxeddu E, Ferretti E, et al. BRAF mutations in papillary thyroid
carcinomas inhibit genes involved in iodine metabolism. J Clin Endocrinol
Metab 2007;92:2840–3.
[41] Nucera C, Lawler J, Parangi S. BRAF(V600E) and microenvironment in thyroid
cancer: a functional link to drive cancer progression. Cancer Res 2011;71:2417–22.
[42] Martin KJ, Fournier MV, Reddy GP, Pardee AB. A need for basic research on
fluid-based early detection biomarkers. Cancer Res 2010;70:5203–6.
[43] Schwarzenbach H, Hoon DS, Pantel K. Cell-free nucleic acids as biomarkers in
cancer patients. Nat Rev Cancer 2011;11:426–37.