Radiology and Oncology  |  Ljubljana  |  Slovenia  |  www.radioloncol.com
Radiol Oncol 2025 doi: 10.2478/raon-2025-0026
1
research article
Expression of the stem cell markers 
NANOG and SOX2 in the cervical squamous 
carcinogenesis
Miha Koren1, Margareta Zlajpah1, Mario Poljak2, Kristina Fujs Komlos2,  
Margareta Strojan Flezar1
1 Institute of Pathology, Faculty of Medicine, University of Ljubljana, Ljubljana, Slovenia
2 Institute of Microbiology and Immunology, Faculty of Medicine, University of Ljubljana, Ljubljana, Slovenia
Radiol Oncol 2025
Received 05 March 2025 
Accepted 26 march 2025
Correspondence to: Assist. Miha Koren, M.D., Institute of Pathology, Faculty of Medicine, University of Ljubljana, Ljubljana, Slovenia.  
E-mail: miha.koren@mf.uni-lj.si 
Disclosure: No potential conflicts of interest were disclosed. 
This is an open access article distributed under the terms of the CC-BY license (https://creativecommons.org/licenses/by/4.0/).
Background. The aim of the present study was to assess a diagnostic potential of stem cell markers NANOG and 
SOX2 for classifying cervical squamous intraepithelial lesions (SILs)/cervical intraepithelial neoplasia (CIN).
Patients and methods. NANOG and SOX2 expression was evaluated immunohistochemically on 40 patients: in 10 
cases each of low-grade SIL (LSIL), high-grade SIL/CIN, grade 2 (HSIL/CIN 2), HSIL/CIN, grade 3 (HSIL/CIN 3), cervical 
squamous cell carcinoma (CSCC) and their adjacent non-dysplastic squamous epithelium. In addition, human papil-
lomavirus (HPV) genotyping and immunohistochemical staining with p16 and Ki-67 were done. NANOG and SOX2 
expression was compared between squamous lesions and controls and between squamous lesions by multiplying 
staining intensity (SI) by the percentage of positive cells (P) and by multiplying SI by the thickness of staining in epithe-
lium (T) to calculate SI x P and SI x T score.
Results. NANOG and SOX2 expression gradually increased from non-dysplastic squamous epithelium via LSIL and 
HSIL to CSCC. Expression of NANOG and SOX2 was higher in LSIL compared to controls (P < 0.05 for NANOG Si x P 
and Si x T scores and SOX2 SI x T score) and lower compared to HSIL (P < 0.05 for all SI x P and SI x T scores). HSIL/CIN 
3 showed higher SOX2 expression than HSIL/CIN 2 (P < 0.05 for SI x P and SI x T scores). 
Conclusions. Contrary to p16, NANOG and SOX2 could be effective for distinguishing LSIL from non-dysplastic 
changes. NANOG and SOX2 could be surrogate markers for differentiating LSIL from HSIL. Moreover, SOX2 could be 
helpful for distinguishing HSIL/CIN 2 from HSIL/CIN 3. Further studies with larger numbers of patients and molecular 
insights are needed.
Key words: cervical cancer; cervical intraepithelial neoplasia; squamous intraepithelial lesion; cancer stem cell; 
NANOG; SOX2
Introduction
Cervical cancer is the fourth most common cancer 
in women worldwide with around 660,000 new 
cases and 350,000 deaths in 2022.1 The highest inci-
dence rates of cervical cancer and over 90% of cer-
vical cancer-related deaths are in low- and middle-
income countries.2 In high-income countries, cer-
vical cancer incidence and mortality have declined 
due to human papillomavirus (HPV) vaccination, 
screening programs and early treatment of precur-
sor lesions.3,4 
Histologically, precursor lesions of the squa-
mous cervical epithelium are classified as squa-
mous intraepithelial lesions (SILs) or cervical in-
traepithelial neoplasia (CIN). They are categorized 
Radiol Oncol 2025
Koren M et al. / Stem cell markers NANOG and SOX2 in the cervical squamous carcinogenesis2
as low-grade SILs (LSILs) encompassing CIN, 
grade 1 (CIN 1), koilocytic atypia and condyloma 
acuminatum, and high-grade SILs (HSILs), based 
on the extension of dysplasia. HSILs may be fur-
ther subdivided into HSIL/CIN, grade 2 (CIN 2) and 
HSIL/CIN, grade 3 (CIN 3).5 Two-tiered LSIL/HSIL 
system is preferred because it reflects improved 
reproducibility and enhanced biological relevance 
compared to three-tiered CIN 1/2/3 system.6 HSIL 
is caused by transforming infection with high-risk 
HPV (HR-HPV) that deregulates expression of E6 
and E7 viral proteins leading to loss of cell cycle 
control and uncontrolled cell proliferation.7,8 
While the majority of LSILs (90%) regress sponta-
neously, about 10% of LSILs subsequently progress 
to HSIL. It is estimated that spontaneous regression 
from HSIL to LSIL occurs in 30% to 50% of cases.5  
HSIL/CIN 2 lesions show even higher potential for 
regression.9 The risk of progression from untreated 
HSIL to cancer is 0.5–1% annually, with approxi-
mately 30% of HSIL/CIN 3 developing into cervical 
cancer over a 30-year period.10 There are currently 
no biomarkers that reliably predict the progression 
of SIL.11,12 Imunohistochemical marker p16 is an ef-
fective diagnostic tool to differentiate HSIL from 
mimickers of precancerous lesions. The use of p16 
immunohistochemistry is also recommended to 
clarify the diagnosis of HISL/CIN 2 versus LSIL in 
morphologically equivocal cases.6 Nevertheless, a 
significant proportion of LSILs show p16 positiv-
ity and despite p16 diagnostic value, its staining 
in LSIL is not predictive for progression of the le-
sion.11,12 Currently, all HSILs are usually surgically 
excised by large loop excision of the transformation 
zone (LLETZ), as it is unable to predict which le-
sion will regress or progress to cervical squamous 
cell carcinoma (CSCC). Since excision of all HSILs 
could lead to overtreatment and unnecessary com-
plications related to woman’s reproductive pros-
pect, new markers are needed to predict the natu-
ral course of precancerous lesions.13 
Recent studies on head and neck squamous 
cell carcinoma (HNSCC) and squamous dysplasia 
have shown a diagnostic and prognostic potential 
of a cancer stem cell (CSC) marker NANOG.14,15 The 
relationship between the expression of CSC mark-
ers and carcinogenesis of various cancers as well 
as CSCC has been studied.16,17 Increasing evidence 
has suggested that tumorigenesis depends on 
CSCs, a small population of cells within a tumour 
that can self-renew and differentiate into multiple 
cell types.16 CSCs may originate from normal stem 
cells and express the same markers (CSC markers) 
as stem cells such as NANOG, OCT3/4, and SOX2.17 
Several studies have shown that the NANOG pro-
tein is highly expressed in cancerous tissues and in 
early embryonic development, whereas its expres-
sion is much lower or absent in normal adult tis-
sues.18 Recent data have demonstrated that SOX2 
expression is related to several human malignan-
cies and it is nowadays considered as a key driver 
of squamous carcinogenesis.19,20 While several 
authors reported SOX2 expression in HSIL/CIN 3 
and CSCC, only few of them have investigated its 
expression in LSIL and HSIL/CIN 2.21-32 Even less 
studies have been published regarding NANOG 
and cervical squamous carcinogenesis.33-36
In previous research of CSCs in HNSCC at our 
institution, more intense NANOG staining was 
observed during progression of head and neck 
dysplasia.37 The aim of the present study was to 
assess the ability of CSC markers to discriminate 
different grades of cervical squamous neopla-
sia. We evaluated the expression of NANOG and 
SOX2 using immunohistochemistry in LSIL, HSIL/
CIN 2, HSIL/CIN 3, CSCC and their adjacent non-
dysplastic squamous epithelium. HPV genotyping 
and immunohistochemical staining with common 
diagnostic markers p16 and Ki-67 were also per-
formed to confirm the accurate classification of the 
lesions.
Patients and methods
Patients
The study was approved by the National Medical 
Ethics Committee, Ministry of Health, Republic 
of Slovenia (consent No. 0120-99/2020/6.) A total 
of 2,478 cases of SILs and CSCCs, obtained dur-
ing routine diagnostic and therapeutic procedures 
from January 2015 to December 2021, were iden-
tified from archives of the Institute of Pathology, 
Medical Faculty, University of Ljubljana, Slovenia. 
After detailed review, 10 samples of each LSIL, 
HSIL/CIN 2, HSIL/CIN 3 and CSCCs from total of 
40 patients with unequivocal diagnoses and suf-
ficient tissue for further analysis were selected. 
Their adjacent non-dysplastic squamous epithe-
lium was used as controls for comparison with 
cervical squamous neoplasia. None of the patients 
involved received chemotherapy, radiotherapy or 
immunotherapy before the specimen collection.
HPV genotyping
All lesions were selected for HPV genotype anal-
ysis. Tissue cores from representative areas of 
Radiol Oncol 2025
Koren M et al. / Stem cell markers NANOG and SOX2 in the cervical squamous carcinogenesis 3
formalin-fixed, paraffin-embedded (FFPE) tissue 
blocks were punched using a 600 µm needle. Total 
DNA from at least three punches per FFPE sam-
ple was extracted by an overnight digestion with 
protease at 55°C, followed by DNA purification us-
ing a MagMax FFPE DNA/RNA Ultra Kit (Applied 
Biosystems; Thermo Fisher Scientific, Waltham, 
MA, USA) according to the manufacturer’s instruc-
tions with one modification: protease digestion 
was performed overnight mixing at 300 rpm for 15 
seconds every 4 minutes instead of 1 hour. DNA 
concentrations were estimated by spectrophoto-
metric analysis at 260 nm using the NanoDropTM 
1000 spectrophotometer (Thermo Fisher Scientific, 
Wilmington, DE, USA). Apart from the deparaffi-
nization solution (xylene; Sigma-Aldrich; Merck 
KGaA, Darmstadt, Germany) and the ethanol 
(Merck KGaA, Darmstadt, Germany), all the rea-
gents were from Applied Biosystems (Thermo 
Fisher Scientific, Waltham, MA, USA).
To determine the presence of HPV, 100 ng of 
DNA was tested with commercial HPV genotyping 
test AnyplexTM II HPV28 Detection (Seegene, Seoul, 
South Korea), which is capable of simultaneous de-
tection and differentiation of 28 HPV genotypes 
(HPV-6, 11, 16, 18, 26, 31, 33, 35, 39, 40, 42, 43, 44, 
45, 51, 52, 53, 54, 56, 58, 59, 61, 66, 68, 69, 70, 73, and 
82).38 Each multiplex PCR reaction was performed 
according to the manufacturer’s instructions.39 
Immunohistochemistry and evaluation
Immunohistochemical analysis of p16, Ki-67, 
NANOG and SOX2 expression was performed 
on 3-4 µm thick FFPE tissue sections. After de-
paraffinization, hydration and automated anti-
gen retrieval (BenchMark ULTRA, Ventana, F. 
Hoffmann-La Roche (Roche Holding AG), Basel, 
Switzerland), slides were incubated with commer-
cially available anti-p16 (Ventana (F. Hoffmann-La 
Roche (Roche Holding AG), Basel, Switzerland), 
cat no. 805-4713, ready to use) anti-ki67 (Agilent 
Dako, Santa Clara, California, USA, cat no. M7240, 
dilution 1:50), anti-NANOG (Cell Signaling 
Technology, Danvers, Massachusetts, USA, cat no. 
4903T, dilution 1 : 100) and anti-SOX2 (Ventana (F. 
Hoffmann-La Roche (Roche Holding AG), Basel, 
Switzerland), cat no. 760-4621, ready to use) anti-
bodies. Reactions were visualized by incubation 
with peroxidase and 3,3′-diaminobenzidine (p16, 
NANOG and SOX2: OptiVIEW DAB Detection Kit, 
Roche, Basel, Switzerland; Ki-67: ultraVIEW, DAB 
Detection Kit, Roche, Basel, Switzerland) and then 
counterstained with hematoxylin. Negative con-
trols omitting the primary antibody binding were 
included in every run of samples. Positive controls 
included HPV-associated oropharyngeal SCC for 
p16, tonsillar lymphatic tissue for Ki-67, testicular 
seminoma, oral SCC and HSIL/CIN 3 for NANOG 
and basal cells of non-dysplastic squamous cervi-
cal epithelium as an internal positive control for 
SOX2. 
All slides were evaluated independently by 
two pathologists (M.S.F. and M.K.) and consen-
sus agreements were reached in discordant cases. 
Expression of p16 was scored as positive (continu-
ous strong nuclear or nuclear plus cytoplasmic 
staining of the basal cell layer with extension up-
ward involving at least one third of the epithelial 
thickness) or negative (any other staining) in dys-
plastic and non-dysplastic squamous epithelium, 
as previously described, and as positive (any posi-
tive staining) or negative for CSCC.6 Expression 
of Ki-67 in the nuclei was assessed by thickness 
of staining in squamous epithelium for SIL and 
controls (0: parabasal cells or no staining, 1: ba-
sal one third, 2: basal two third, 3: full thickness) 
and for CSCC semiquantitatively according to the 
percentage of positive cells (0: 0%, 1: 1% to 29%, 2: 
30% to 59%, 3: 60% to 100%).29,40 In all controls, SIL 
and CSCC, the expression of NANOG and SOX2 
was assessed by the percentage of positive cells, 
staining intensity and staining pattern (nuclear, 
cytoplasmic, nuclear or cytoplasmic); SOX2 stain-
ing intensity and percentage of positive cells were 
evaluated above the basal layer. Staining intensity 
of NANOG (0: no staining, 1: weak, 2: moderate, 3: 
strong - as in testicular seminoma on slide control) 
and SOX2 (0: no staining, 1: weak, 2: moderate - 
as basal cells of adjacent non-dysplastic squamous 
epithelium, 3: strong) was scored in a maximally 
stained area. Based on previously published stud-
ies, the percentage of positive cells was evaluated 
semiquantitatively and divided into five categories 
for both NANOG (0: < 5%, 1: 5-25%, 2: 25-50%, 3: 
50-75%, 4 > 75%) and SOX2 (0: 0-10%, 1: 10 -25%, 
2: 25-50%, 3: 50-75%, 4 > 75%).21,24,34,37 In addition, 
the expression of NANOG (0: no staining, 1: ba-
sal one third, 2: basal two third, 3: full thickness) 
and SOX2 (0: basal cells or no staining, 1: basal 
one third, 2: basal two third, 3: full thickness) in 
SIL and controls was assessed based on the thick-
ness of staining in squamous epithelium.29 For 
both NANOG and SOX2, staining intensity and 
percentage of positive cells (SIxP score) as well as 
staining intensity and thickness of staining (SIxT 
score) were multiplied to calculate SIxP score and 
SIxT score.21,24,34
Radiol Oncol 2025
Koren M et al. / Stem cell markers NANOG and SOX2 in the cervical squamous carcinogenesis4
Statistical analysis
Statistical analysis was performed using IBM SPSS 
Statistics 24.0 software (IBM Corp.). The Mann-
Whitney U test was used and differences were 
considered as statistically significant at cut-off p ≤ 
0.05 (2-tailed).
Results 
Patients
The mean age of patients in the LSIL group was 
47 years, 37 years in the HSIL/CIN 2 group, 41.1 
years in the HSIL/CIN 3 group and 47.6 years in 
the CSCC group. One out of 10 LSILs was obtained 
by hysterectomy, all other SILs were excised by 
cone biopsy (9 cases) or LLETZ (20 cases). In the 
CSCC group, 1 of 10 samples was a hysterectomy, 
2 of 10 were cervical biopsies, 2 of 10 were LLETZs 
and 5 of 10 were cone biopsies. In 37 of 40 cases, 
adjacent non-dysplastic squamous epithelium un-
der hormonal stimulation (normal reproductive 
age ectocervical epithelium) was found and used 
as control. In a case of HSIL/CIN 2 and in a case of 
HSIL/CIN 3, adjacent squamous metaplastic epi-
thelium was used as their controls and in a case 
FIGURE 1. Representative images of p16 (b, g, l, q, v), Ki-67 (c, h, m r, w), NANOG (d, i, n, s, x) and SOX2 (e, j, o, t, y) immunohistochemical staining in 
a control (a-e), low-grade squamous intraepithelial lesion (LSIL) (f-j), high-grade squamous intraepithelial lesion/cervical intraepithelial neoplasia 
grade 2 (HSIL/CIN 2) (k-o), high-grade squamous intraepithelial lesion/cervical intraepithelial neoplasia grade 3 (HSIL/CIN 3) (p-t) and cervical 
squamous cell carcinoma (CSCC) (u-y). 
HE = haematoxylin and eosin stain
Radiol Oncol 2025
Koren M et al. / Stem cell markers NANOG and SOX2 in the cervical squamous carcinogenesis 5
of CSCC adjacent atrophic squamous epithelium 
was used as a control.
HPV genotyping
The presence of HPV DNA was tested and con-
firmed in all 40 patients included in the study. 
Distribution of HPV genotypes according to histo-
logical diagnosis is shown in Table 1. Single HPV 
genotype infection was found in 1 of 10 LSILs, 7 
of 10 HSIL/CIN 2 cases, 8 of 10 HSIL/CIN 3 cases 
and 6 of 10 CSCCs. Coinfection with two or more 
HPV genotypes was found in 9 of 10 LSILs, 3 of 10 
HSIL/CIN 2 cases, 2 of 10 HSIL/CIN 3 cases and 
4 of 10 CSCCs. In 2 LSILs and 1 HSIL/CIN 2 cas-
es, only possibly carcinogenic HPV genotypes 66 
and 70 and no HR-HPV genotypes were present. 
The remaining 37 patients were infected with HR-
HPV genotypes 16, 18, 31, 39, 51, 52, 56, 58 and 59.7,41 
HPV-16 was the most common genotype, followed 
by HPV-31, HPV-51 and HPV-66.
Immunohistochemistry
Immunohistochemical analysis of p16, Ki-67, 
NANOG and SOX2 was performed for all selected 
samples. Representative images for each group are 
presented in Figure 1. All 40 included control tis-
sues were immunohistochemically negative for 
p16 and expression of Ki-67 was limited on the 
parabasal layer. Two of 10 LSILs were positive for 
p16 and 2 of 10 LSILs were negative for p16. Six of 
10 LSILs were negative for p16 in the major part of 
the lesion and positive for p16 in the minor part of 
the lesion. All 20 cases of HSIL and all 10 cases of 
CSCC were positive for p16 and showed full thick-
ness and/or overall expression of Ki-67. In LSIL, Ki-
67 expression was in full thickness of epithelium 
in 8 out of 10 cases and predominantly in the basal 
two thirds of epithelium in 2 out of 10 cases.
Immunohistochemical staining for NANOG 
was present in the cytoplasm of squamous cells; 
no nuclear staining was observed. In general, con-
TABLE 1. Results of human papillomavirus (HPV) genotyping in squamous lesions
HPV
type
HPV 
16
HPV 
18
HPV 
31
HPV 
39
HPV 
51
HPV 
52
HPV 
53
HPV 
56
HPV 
58
HPV 
59
HPV 
68
HPV 
66
HPV 
70
HPV 
42
LSIL, n 4 0 0 3 2 0 3 1 0 1 2 5 2 2
HSIL/CIN 2, n 4 1 1 0 1 2 0 1 2 0 0 1 0 0
HSIL/ CIN 3, n 8 0 3 0 0 0 1 0 0 1 0 0 0 0
CSCC, n 6 2 2 0 3 0 0 0 0 0 0 0 1 0
CSCC = cervical squamous cell carcinoma; HPV = human papillomavirus; HSIL/CIN 2 = high-grade squamous intraepithelial lesion/cervical intraepithelial neoplasia grade 
2; HSIL/CIN 3 = high-grade squamous intraepithelial lesion/cervical intraepithelial neoplasia grade 3; LSIL = low-grade squamous intraepithelial lesion; n = number of cases
TABLE 2. Evaluation of NANOG staining in controls and squamous lesions
NANOG SI, 
n (%)
NANOG P, 
n (%)
NANOG T, 
n (%)
Score 0 1 2 3 0 1 2 3 4 0 1 2 3
Controls 
(n = 40)
3 
(7.5)
32 
(80)
5 
(12.5)
0 
(0)
3 
(7.5)
6 
(15)
22 
(55)
8 
(20)
1 
(2.5)
3 
(7.5)
4 
(10)
28 
(70)
5 
(12.5)
LSIL 
(n = 10)
0 
(0)
2 
(20)
7 
(70)
1 
(10)
0 
(0)
0 
(0)
0 
(0)
1 
(10)
9 
(90)
0 
(0)
0 
(0)
0 
(0)
10 
(100)
HSIL/CIN 2 
(n = 10)
1 
(10)
0 
(0)
4 
(40)
5 
(50)
1
(10)
0 
(0)
0 
(0)
0 
(0)
9 
(90)
1 
(0)
0 
(0)
0 
(0)
9 
(90)
HSIL/CIN 3 
(n = 10)
0 
(0)
0 
(0)
1 
(10)
9 
(90)
0 
(0)
0 
(0)
0 
(0)
0 
(0)
10 
(10)
0 
(0)
0 
(0)
0 
(0)
10 
(100)
CSCC 
(n = 10)
0 
(0)
0 
(0)
1 
(10)
9 
(90)
0 
(0)
0 
(0)
0 
(0)
0 
(0)
10 
(100) Not applicable
1
CSCC = cervical squamous cell carcinoma; HSIL/CIN 2 = high-grade squamous intraepithelial lesion/cervical intraepithelial neoplasia grade 2; HSIL/CIN 3 = high-grade 
squamous intraepithelial lesion/cervical intraepithelial neoplasia grade 3; LSIL = low-grade squamous intraepithelial lesion; n = number of cases; P = percentage of positive 
cells; SI = staining intensity for NANOG; T = thickness of staining; % = percentage of all cases in each category
1 T was not evaluated in CSCC.
Radiol Oncol 2025
Koren M et al. / Stem cell markers NANOG and SOX2 in the cervical squamous carcinogenesis6
TABLE 3. Evaluation of SOX2 staining in controls and squamous lesions
SOX2 SI, 
n (%)
SOX2 P, 
n (%)
SOX2 T, 
n (%)
Score 0 1 2 3 0 1 2 3 4 0 1 2 3
Controls
(n = 40)
2 
(5)
33 
(82.5)
5 
(12.5)
0 
(0)
2 
(5)
1 
(2.5)
26 
(65)
10 
(25) 0 (0)
2 
(5)
1 
(2.5)
31 
(77.5)
6 
(15)
LSIL
(n = 10)
0 
(0)
4 
(40)
6 
(60)
0 
(0)
0 
(0)
0 
(0)
2 
(20)
4 
(40)
4 
(40)
0 
(0)
0 
(0)
4 
(40)
6 
(60)
HSIL/CIN 2
(n = 10)
0 
(0)
1 
(10)
7 
(70)
2 
(20)
0 
(0)
0 
(0)
0 
(0)
4 
(40)
6 
(60)
0 
(0)
0 
(0)
1 
(10)
9 
(90)
HSIL/CIN 3
(n = 10)
0 
(0)
1 
(10)
1 
(10)
8 
(80)
1 
(10)
0 
(0)
0 
(0)
1 
(10)
8 
(80)
0 
(0)
0 
(0)
0 
(0)
10 
(100)
CSCC
(n = 10)
0 
(0)
0 
(0)
3 
(30)
7 
(70)
0 
(0)
0 
(0)
0 
(0)
2 
(20)
8 
(80) Not applicable
1
CSCC = cervical squamous cell carcinoma; HSIL/CIN 2 = high-grade squamous intraepithelial lesion/cervical intraepithelial neoplasia grade 2; HSIL/CIN 3 = high-grade 
squamous intraepithelial lesion/cervical intraepithelial neoplasia grade 3; LSIL = low-grade squamous intraepithelial lesion; n = number of cases; P = percentage of positive 
cells; SI = staining intensity for NANOG; T = thickness of staining; % = percentage of all cases in each category
1 T was not evaluated in CSCC.
TABLE 4. Comparison of NANOG and SOX2 staining between squamous lesions and their adjacent control non-dysplastic squamous epithelium
NANOG SIxP score,
a ± SD p-value
1 SOX2 SIxP score,
a ± SD p-value
1
C (LSIL) LSIL C (LSIL) LSIL
2.30 ± 0.95 7.50 ± 2.46 0.000 3.00 ± 1.41 5.40 ± 2.63 0.054
C (HSIL/CIN 2) HSIL/CIN 2 C (HSIL/CIN 2) HSIL/CIN 2
3.30 ± 2.06 9.20 ± 3.79 0.002 2.20 ± 1.48 7.70 ± 2.75 0.000
C (HSIL/CIN 3) HSIL/CIN 3 C (HSIL/CIN 3) HSIL/CIN 3
2.00 ± 0.94 10.40 ± 2.07 0.000 2.30 ± 1.16 10.10 ± 3.84 0.002
C (CSCC) CSCC C (CSCC) CSCC
1.44 ± 1,01 11.56 ± 1.33 0.000 2.44 ± 0,53 10.22 ± 2.73 0.000
C (HSIL) HSIL C (HSIL) HSIL
2.65 ± 1.65 9.80 ± 2.96 0.000 2.25 ± 1.26 8.90 ± 3.39 0.000
NANOG SIxT score2
a ± SD p-value
1 SOX2 SIxT score2,
a ± SD p-value
1
C (LSIL) LSIL C (LSIL) LSIL
1.90 ± 0.32 5.70 ± 1.70 0.000 2.50 ± 0.85 4.30 ± 1.89 0.0027
C (HSIL/CIN 2) HSIL/CIN 2 C (HSIL/CIN 2) HSIL/CIN 2
3.00 ± 1.76 6.90 ± 2.85 0.005 1.90 ± 0.99 6.10 ± 1.85 0.000
C (HSIL/CIN 3) HSIL/CIN 3 C (HSIL/CIN 3) HSIL/CIN 3
2.20 ± 1.48 7.80 ± 1.55 0.000 2.40 ± 1.17 8.10 ± 2.02 0.000
C (HSIL) HSIL C (HSIL) HSIL
2.60 ± 1.59 7.35 ± 2.22 0.000 2.15 ± 1.06 7.10 ± 2.10 0.000
a = average; C (HSIL) = control non-dysplastic squamous epithelium adjacent to high-grade squamous intraepithelial lesion; C (LSIL) = control non-dysplastic squamous 
epithelium adjacent to low-grade squamous intraepithelial lesion; C (HSIL/CIN 2) = control non-dysplastic squamous epithelium adjacent to high-grade squamous 
intraepithelial lesion/cervical intraepithelial neoplasia grade 2; C (HSIL/CIN 3) = control non-dysplastic squamous epithelium adjacent to high-grade squamous 
intraepithelial lesion/cervical intraepithelial neoplasia grade 3; CSCC = cervical squamous cell carcinoma; C (CSCC) = control non-dysplastic squamous epithelium 
adjacent to cervical squamous cell carcinoma; HSIL = high-grade squamous intraepithelial lesion; HSIL/CIN 2 = high-grade squamous intraepithelial lesion/cervical 
intraepithelial neoplasia grade 2; HSIL/CIN 3 = high-grade squamous intraepithelial lesion/cervical intraepithelial neoplasia grade 3; LSIL = low-grade squamous 
intraepithelial lesion; SD = standard deviation; SIxP score = staining intensity (SI) multiplied by percentage of positive cells (P); SIxT score = staining intensity (SI) multiplied 
by thickness of staining (T)
1 Comparisons between groups were tested by the Mann-Whitney test.
2 SIxT score was not calculated for CSCC because T was not evaluated in CSCC.
Radiol Oncol 2025
Koren M et al. / Stem cell markers NANOG and SOX2 in the cervical squamous carcinogenesis 7
trols showed lower staining intensity, percentage 
and thickness of staining of NANOG than their 
adjacent SIL and CSCC (Table 2). No reaction was 
observed in 3 of 40 (7.5 %) controls; one of them 
was an atrophic squamous epithelium and two 
were normal reproductive age ectocervical epi-
thelium. In other controls, staining intensity was 
mostly weak (in 80 %) and extending up to basal 
two thirds of epithelium (in 87.5%), with less than 
50% of cells being positive in 31 of 40 controls. In 
a majority of LSILs, staining intensity was moder-
ate (in 70%) and positive in more than 75% of cells 
(in 90%), with full thickness expression observed 
in all cases. Moderate or strong staining intensity 
was found in 9 out of 10 cases of HSIL/CIN 2 and 
in all 10 cases of HSIL/CIN 3. More than 75% of 
positive cells and full thickness staining was ob-
served in the majority of HSIL/CIN 2 (in 90%) and 
in all HSIL/CIN 3. The reaction was mostly strong 
(in 90%) and in more than 75% of cells in all cases 
of CSCC.
Immunohistochemical staining for SOX2 was 
observed only in the nuclei of squamous cells. 
SIL and CSCC showed higher staining inten-
sity, percentage and thickness of SOX2 staining 
than their adjacent normal squamous epithelium 
(Table 3). Moderate intensity of staining in the ba-
sal cells was present in 37 of 40 controls (92.5%). 
Weak SOX2 reaction was observed in basal and 
upper cells of an atrophic squamous epithelium. 
Complete absence of SOX2 staining was in the two 
cases of control squamous metaplastic epithelium. 
Staining intensity was mostly weak (in 82.5%), 
with less than 50% positive cells (in 72.5%) and ex-
tending up to basal two thirds of epithelium (in 
85%) in other controls. Majority of LSILs showed 
moderate staining intensity (in 60%) in more than 
50% of cells (in 80%), extending at least up to basal 
two thirds of epithelium in all cases. In HSIL/CIN 
2, moderate staining intensity (in 70%), involving 
full thickness of epithelium (in 90%), was mostly 
observed, with more than half of positive cells 
TABLE 5. Comparison of NANOG and SOX2 staining between squamous lesions
NANOG SIxP score,
a ± SD p-value
1 SOX2 SIxP score, 
a ± SD p-value
1
LSIL HSIL/CIN 2 LSIL HSIL/CIN 2
7.50 ± 2.46 9.20 ± 3.79 0.101 5.40 ± 2.63 7.70 ± 2.75 0.114
HSIL/CIN 2 HSIL/CIN 3 HSIL/CIN 2 HSIL/CIN 3
9.20 ± 3.79 10.40 ± 2.07 0.547 7.70 ± 2.75 10.10 ± 3.84 0.031
HSIL/CIN 3 CSCC HSIL/CIN 3 CSCC
10.40 ± 2.07 11.56 ± 1.33 0.165 10.10 ± 3.84 10.22 ± 2.73 0.805
LSIL HSIL LSIL HSIL
7.50 ± 2.46 9.80 ± 2.96 0.018 5.40 ± 2.63 8.90 ± 3.39 0.009
HSIL CSCC HSIL CSCC
9.80 ± 2.96 11.56 ± 1.33 0.078 8.90 ± 3.39 10.22 ± 2.73 0.370
NANOG SIxT score2
a ± SD p-value
1 SOX2 SIxT score2,
a ± SD p-value
1
LSIL HSIL/CIN 2 LSIL HSIL/CIN 2
5.70 ± 1.70 6.90 ± 2.85 0.101 4.30 ± 1.89 6.10 ± 1.85 0.056
HSIL/CIN 2 HSIL/CIN 3 HSIL/CIN 2 HSIL/CIN 3
6.90 ± 2.85 7.80 ± 1.55 0.547 6.10 ± 1.85 8.10 ± 2.02 0.023
LSIL HSIL LSIL HSIL
5.70 ± 1.70 7.35 ± 2.22 0.018 4.30 ± 1.89 7.10 ± 2.10 0.003
a = average; CSCC = cervical squamous cell carcinoma; HSIL = high-grade squamous intraepithelial lesion; HSIL/CIN 2 = high-grade squamous intraepithelial lesion/
cervical intraepithelial neoplasia grade 2; HSIL/CIN 3 = high-grade squamous intraepithelial lesion/cervical intraepithelial neoplasia grade 3; LSIL = low-grade squamous 
intraepithelial lesion; SD = standard deviation; SIxP score = staining intensity (SI) multiplied by percentage of positive cells (P); SIxT score = staining intensity (SI) multiplied 
by thickness of staining (T)
1 Comparisons between groups were tested by the Mann-Whitney test.
2 SIxT score was not calculated for CSCC because T was not evaluated in CSCC.
Radiol Oncol 2025
Koren M et al. / Stem cell markers NANOG and SOX2 in the cervical squamous carcinogenesis8
in all cases. Majority of HSIL/CIN 3 (in 80%) and 
CSCC (in 70%) strongly expressed SOX2 in more 
than 75% of cells (in 80%). In all HSIL/CIN 3, full 
thickness of epithelium was involved.
Comparison between non-dysplastic 
squamous epithelia and squamous 
lesions
Statistical analysis of SIxP score and SIxT score for 
NANOG and SOX2 for controls, LSIL, HSIL/CIN 2, 
HSIL/CIN 3 and CSCC was performed and results 
are summarized in Table 4. Both NANOG scores 
were statistically significantly higher in LSIL, 
HSIL/CIN 2, HSIL/CIN 3 and CSCC compared to 
their adjacent non-dysplastic squamous epitheli-
um. LSIL, HSIL/CIN 2, HSIL/CIN 3 and CSCC also 
showed higher scores for SOX2 than their adjacent 
control epithelium. The difference was statistically 
significant in all cases except when LSIL SIxP score 
was compared to their controls. However, even in 
this case, it was very close to the cut-off value of p 
≤ 0.05 (p = 0.054).
Comparison between squamous lesions
Comparison of SIxP score and SIxT score for 
NANOG and SOX2 between lesions was statisti-
cally analysed and results are presented in Table 5. 
NANOG scores were higher in CSCC compared to 
HSIL/CIN 3 and HSIL, in HSIL/CIN 3 compared to 
HSIL/CIN 2, in HSIL/CIN 2 compared to LSIL and 
in HSIL compared to LSIL. However, the difference 
was statistically significant only between HSIL 
and LSIL. SOX2 scores were also higher in CSCC 
compared to HSIL/CIN 3 and HSIL, in HSIL/CIN 3 
compared to HSIL/CIN 2, in HSIL/CIN 2 compared 
to LSIL and in HSIL compared to LSIL. Statistically 
significant difference was between HSIL and LSIL 
and, interestingly, also between HSIL/CIN 3 and 
HSIL/CIN 2.
Discussion
According to the current knowledge, CSCs are 
the key factor in tumour initiation and develop-
ment, metastasis, and recurrence.16 CSCs and nor-
mal stem cells share similar signalling pathways 
and transcription factors, including NANOG and 
SOX2.17 These proteins have been found over-
expressed in cells of variety of precancerous 
and cancerous lesions, exhibiting an important 
function during carcinogenesis. By expressing 
NANOG and SOX2 among the others, these cells 
show CSC-like properties.42,43 In the present study, 
we demonstrated the increasing expression of two 
common markers, NANOG and SOX2, during cer-
vical squamous carcinogenesis and we confirmed 
their role in the progression of SIL to CSCC.
The most prevalent HPV type 16 in women 
with normal cervical cytology, precancerous cervi-
cal lesions and invasive cervical cancer was also 
the most common HPV genotype in our study.44 
Infection with at least one HPV genotype was con-
firmed in all patients, and at least one HR-HPV 
genotype was present in 37 of 40 (92.5%) patients. 
As coinfection with more than one HPV genotype 
is a common event, we detected multiple HPV 
types in 45% of all patients, as expected.45,46 The 
majority of cervical SCC are HPV-related and de-
velop from HSIL as a consequence of uncontrolled 
expression of the HR-HPV viral oncogenes E6 and 
E7.5,7 Recently, HR-HPV E6 and E7 oncoproteins 
have been reported to increase the expression of 
stem cell genes OCT3/4, NANOG and SOX2 and 
promote cell self-renewal upon inactivation of tu-
mour suppressor proteins p53 and Rb.47-51 Other 
studies described a regulatory effect of NANOG 
and SOX2 on the expression of HR-HPV E6 and 
E7 oncogenes, so research regarding CSC markers 
and HPV infection is needed.52,53 Unfortunately, 
due to the small number of cases without HR-HPV 
infection, further analysis of NANOG and SOX2 
expression between HR-HPV and non-HR-HPV 
group was not feasible in our present work.
NANOG is expressed in a variety of human ma-
lignancies, particularly breast cancer, colon cancer, 
head and neck cancer, lung cancer and pancreatic 
cancer. High levels of NANOG have been associat-
ed with metastatic disease and poor prognosis.14,18 
While several studies have been conducted on 
NANOG in various organs, there are limited data 
on its expression in cervical squamous neoplasia.
In the present study, we observed cytoplasmic 
pattern of NANOG staining in the cervical squa-
mous epithelium and associated neoplastic chang-
es, whereas previous studies in the field reported 
cytoplasmic and/or nuclear expression.33,34,36 In 
our research group, predominantly cytoplasmic 
NANOG staining was observed in mild squamous 
dysplasia (very weak staining intensity) and in se-
vere dysplasia (strong staining intensity) in head 
and neck squamous neoplasia.37 However, as a 
transcription factor, NANOG would be expected 
to be localized in the nucleus to function and ac-
cordingly, its nuclear pattern of expression has 
been observed in germ cell tumours.37,54 In other 
Radiol Oncol 2025
Koren M et al. / Stem cell markers NANOG and SOX2 in the cervical squamous carcinogenesis 9
organs, NANOG has been detected mainly in the 
cytoplasm or mainly in the nuclei of malignant 
tumours, however some studies demonstrated 
NANOG in both sites of the cell.18 Protein modi-
fication and/or spatial structure changes could be 
one of the factors involved in cellular translocation 
and different distribution patterns of NANOG 
protein and different staining patterns conse-
quetally.34 The cellular translocation of NANOG 
protein from nucleus to cytoplasm could be relat-
ed to the molecular characteristics of the human 
NANOG protein, which has a region for nuclear 
localization and another region for nuclear export, 
suggesting a shuttling mechanism between nucle-
us and cytoplasm.55 Moreover, different staining 
patterns of NANOG could be furtherly explained 
by an antibody-dependent mechanism. Different 
antibodies were used in studies and some of them 
could possibly bind to the nuclear NANOG pro-
tein and some to the exported NANOG protein, or 
both.34,56
We found higher expression of NANOG in all 
SIL categories and CSCC compared to adjacent non-
dysplastic squamous epithelium. We also found 
higher NANOG expression in HSIL compared to 
LSIL, however we did not observe any significant 
difference in NANOG expression between HSIL 
and CSCC (p-value = 0.078). Noh et al. previously 
reported increased expression of NANOG in HSIL 
compared to LSIL and in CSCC compared to HSIL. 
Similar to our findings, they also described higher 
expression of NANOG protein in HSIL and CSCC 
compared to normal cervical epithelium.35 Ye et al. 
also, demonstrated higher expression of NANOG 
in LSIL compared to normal cervical squamous ep-
ithelium, in HSIL compared to LSIL, and in CSCC 
compared to HSIL.36 The small number of cases in 
our study may explain that we did not find a statis-
tically significant difference in NANOG staining 
between HSIL and CSCC.
SOX2 is a key transcription factor that is ex-
pressed during embryonic development.16,57 
Amplification of the SOX2 gene locus, leading to 
an increased expression, is supposed to be an im-
portant factor in cancerogenesis, especially in the 
pathogenesis of squamous cell carcinomas of vari-
ous sites.19,20 
In our study, SOX2 expression was localised in 
the nuclei of cervical squamous epithelium and 
associated neoplastic changes, as previously re-
ported.19-32 In the non-dysplastic (non-atrophic and 
non-metaplastic) cervical squamous epithelium, 
we observed a consistent, moderate intensity of 
positive staining in basal cells, concordant with 
previous reports.19,21,23,26,30 Therefore, we consid-
ered SOX2 expression in the upper layers of squa-
mous epithelium as abnormal and evaluated the 
intensity of SOX2 staining and the percentage of 
positive cells only above the basal layer.
We found increased expression of SOX2 in all 
grades of SIL and in CSCC, which is consistent 
with the findings by Kim et al.28 However, most 
other studies were mainly focused on the expres-
sion of SOX2 in CSCC and/or HSIL/CIN 3 and were 
reporting its increased expression in these neo-
plastic lesions.19,21-24 In our study, we observed an 
increase in SOX2 expression from LSIL to HSIL. In 
addition, we showed a significantly higher expres-
sion of SOX2 in HSIL/CIN 3 compared to HSIL/
CIN 2. Kim et al. reported a gradual increase in 
SOX2 expression from normal cervical squamous 
epithelium to LSIL and HSIL, but no comparison 
was made between HSIL/CIN 2 and HSIL/CIN 3 
categories.28 Additionally, they reported higher 
SOX2 expression in CSCC than in HSIL, which 
was also confirmed by Atigan et al.28,31 However, 
in our study, we found no difference in SOX2 ex-
pression between HSIL/CIN 3 or HSIL and CSCC, 
which is consistent with some other studies.22,23,29 
Later research encompassing cervical squamous 
dysplasia and CSCC confirmed higher SOX2 ex-
pression in HSIL than in LSIL. A higher percent-
age of SOX2 nuclear staining in HSIL than in 
LSIL was described by Atigan et al.31 Wolsky et al. 
demonstrated statistically significant difference in 
SOX2 distribution between LSIL and HSIL, with 
similar results displayed by p16 and Ki-67. They 
even suggested that SOX2 is a diagnostic marker 
comparable to p16 and Ki-67 for distinguishing be-
tween LSIL and HSIL.29 
The current WHO recommendation for cervi-
cal cytological or tissue specimens is to use a two-
tiered classification of squamous lesions, namely 
LSIL and HSIL terminology reflecting HPV-driven 
pathogenesis and establishing improved repro-
ducibility of the LSIL/HSIL classification system.5,6 
In HSIL category, lesions corresponding to HSIL/
CIN 2, can be difficult to distinguish from LSIL. 
Block-type positive p16 staining supports the cate-
gorization of HSIL/CIN 2, while diagnosis of HSIL/
CIN 2 is unlikely in the absence of p16 block-type 
positivity.6 On the other hand, up to 50% of LSILs 
could be block-type p16 positive.11 In the present 
study, we observed p16 positivity in the entire le-
sion in 2 of 10 LSILs and focal block-type p16 posi-
tivity next to heterogeneous or negative staining in 
6 of 10 LSILs. In unequivocal LSILs, such p16 posi-
tivity is not conclusive for HSIL/CIN 2.6,11,12 Results 
Radiol Oncol 2025
Koren M et al. / Stem cell markers NANOG and SOX2 in the cervical squamous carcinogenesis10
of our study suggest NANOG and SOX2 as sup-
portive diagnostic markers for the differentiation 
between LSIL and HSIL, which was proposed in 
some previously published studies.28,29,31,35,36
Higher expression of NANOG and SOX2 in 
LSIL compared to non-dysplastic cervical squa-
mous epithelium could also help to distinguish 
between reactive changes and LSIL.28,36 Reactive 
or inflammatory features of non-dysplastic cervi-
cal squamous epithelium can mimic LSIL, result-
ing in overdiagnosis of LSIL in this context.58,59 The 
biomarker p16 has no utility at this diagnostic in-
terface, as many LSILs are p16-negative. Moreover, 
the use of p16 is not recommended to diagnose 
LSILs with typical morphology.6 Therefore, past 
research and our data offer an opportunity to 
NANOG and SOX2 in that field. Areas of im-
mature squamous metaplasia were previously 
described unstained with SOX2, and in our case 
of atrophic squamous epithelium there was no 
NANOG staining.26,30 Nevertheless, the significant 
difference in NANOG and SOX2 (SIxT score) ex-
pression between LSIL and non-dysplastic cervical 
squamous epithelium found in our study predict 
a potential of those two markers to identify LSIL. 
However, as it is a common diagnostic problem to 
distinguish between reactive changes and LSIL, 
further studies that include reactive or inflamed 
cervical squamous epithelium are needed.6,58,59 
In general, all HSILs are treated by excision, as 
there are currently no reliable biomarkers for pre-
dicting the progression or regression of HSIL.11-13 
However, it is known that HSIL/CIN 2 has a sig-
nificantly higher regression rate than HSIL/CIN 
3.9 Therefore, young women who wish to pre-
serve fertility could be treated less aggressively.5,40 
Accurate differentiation between HSIL/CIN 2 and 
HSIL/CIN 3 can be challenging and p16 immu-
nostaining is not helpful as it is positive for both, 
HSIL/CIN 2 and HSIL/CIN 3 in most cases.6,58 In 
our study, we found a significantly higher SOX2 
expression in HSIL/CIN 3 compared to HSIL/CIN 
2, indicating a diagnostic potential of SOX2 as a bi-
omarker in diagnostically challenging cases in this 
context. A specific expression pattern of SOX2 for 
HSIL/CIN 3 was also described by Moshi et al., pro-
viding a possibility for SOX2 to identify HSIL/CIN 
3. Unfortunately, the distribution pattern of SOX2 
staining throughout SIL in their study was some-
times complex, with a mixed and discontinuous 
pattern, consequentially limiting the use of SOX2 
for recognizing HSIL/CIN 3.32 However, studies 
including more cases are needed to further assess 
the value of CSC markers in routine practice.
The present study has several strengths and 
limitations. The main strength is our systematic 
approach and the inclusion of the comparable 
number of cases in all SIL categories and CSCC. 
This enabled insight into NANOG and SOX2 ex-
pression during the whole process of cancerogen-
esis. Another strength is the fact that in addition 
to staining intensity and percentage of positive 
cells, we also evaluated the thickness of staining 
in squamous epithelium for NANOG and SOX2. 
We believe that immunohistochemical markers 
regarding cervical SIL should be assessed by the 
thickness of staining in the epithelium. Our belief 
is supported by at least two facts. First, the cur-
rent WHO categorization of SIL is based on the 
thickness of dysplastic changes in epithelium.5 
In addition, several previous studies have evalu-
ated different markers according to the thickness 
of staining in squamous epithelium and specifi-
cally, p16 positivity in cervical SIL is defined as 
continuous staining in at least basal third of the 
epithelium.6,29,40 Another strength of our study is 
objective evaluation system for both NANOG and 
SOX2. We used the same criteria for the percent-
age of positive cells as some previously published 
studies, allowing a good comparison.21,24,34,37 We al-
so established strict criteria for NANOG and SOX2 
staining intensity, which were mostly undefined 
in previous studies. In contrast, the personal deci-
sion for assessment of NANOG and SOX2 staining 
intensity in a maximally stained area is the main 
limitation of our study. This could potentially over-
estimate the immunohistochemical expression of 
both markers. However, scoring in a maximally 
stained area is less subjective than scoring of the 
entire lesion. The small number of involved cases 
also limits the interpretations of our results. We 
only focused on immunohistochemical analysis of 
NANOG and SOX2 expression, which is another 
limitation. Further molecular studies of NANOG, 
SOX2 and their regulation need to be performed 
to understand their exact role in cervical carcino-
genesis.
Conclusions
We have shown an increased NANOG and SOX2 
expression in LSIL, HSIL and CSCC compared to 
non-dysplastic cervical squamous epithelium. Our 
findings confirm the potential of NANOG and 
SOX2 in distinguishing LSIL from HSIL as sur-
rogates for p16. NANOG and SOX2, in contrast to 
p16, could be suitable for the identification of LSIL 
Radiol Oncol 2025
Koren M et al. / Stem cell markers NANOG and SOX2 in the cervical squamous carcinogenesis 11
in a background of cervical squamous epithelium 
exhibiting non-neoplastic changes mimicking SIL/
CIN. In addition, SOX2 could be a marker for dif-
ferentiation of HSIL/CIN 2 from HSIL/CIN 3. A 
clear advantage of SOX2 compared to NANOG 
and p16 is its uniform and easily interpretable nu-
clear staining. Further studies with larger num-
bers of patients and molecular insights should fo-
cus on the diagnostic and prognostic significance 
of NANOG and SOX2 in the entire process of cer-
vical carcinogenesis.
Acknowledgments 
This study was funded by the Slovenian Research 
and Innovation Agency (research core Fundings 
No. P3-0054 and P3-0083).
Author contribution 
Study concepts: MK, MP, MSF. Study design: MK, 
MP, MSF. Case selection: MK, MSF. Slide evalua-
tion: MK, MSF. HPV genotyping: KFK, MP, MZ. 
Data analysis: MK, MŽ. Data interpretation: KFK, 
MK, MP, MSF, MZ. Manuscript preparation: MK, 
MSF, MZ. Manuscript editing: KFK, MK, MP, MSF, 
MZ. Manuscript review: KFK, MK, MP, MSF, MZ.
References
1. Bray F, Laversanne M, Sung H, Ferlay J, Siegel RL, Soerjomataram I, et al. 
Global cancer statistics 2022: GLOBOCAN estimates of incidence and mor-
tality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2024; 74: 
229-63. doi: 10.3322/caac.21834
2. World Health Organisation. Cervical cancer [Internet]. Geneva: World 
Health Organisation; 2024. [cited 2025 Jan 10]. Available at: https://www.
who.int/news-room/fact-sheets/detail/cervical-cancer
3. Jemal A, Simard EP, Dorell C, Noone AM, Markowitz LE, Kohler B, et al. 
Annual report to the nation on the status of cancer, 1975-2009, featuring 
the burden and trends in human papillomavirus(HPV)-associated cancers 
and HPV vaccination coverage levels. J Natl Cancer Inst 2013; 105: 175-201. 
doi: 10.1093/jnci/djs491
4. de Sanjose S, Brotons M, LaMontagne DS, Bruni L. Human papillomavirus 
vaccine disease impact beyond expectations. Curr Opin Virol 2019; 39: 16-
22. doi: 10.1016/j.coviro.2019.06.006
5. WHO Classification of Tumours Editorial Board. Female genital tumours. 5th 
edition. Lyon: International Agency for Research on Cancer; 2020.
6. Darragh TM, Colgan TJ, Thomas Cox J, Heller DS, Henry MR, Luff RD, et 
al; Members of the LAST Project Work Groups. The Lower Anogenital 
Squamous Terminology Standardization project for HPV-associated le-
sions: background and consensus recommendations from the College 
of American Pathologists and the American Society for Colposcopy and 
Cervical Pathology. Int J Gynecol Pathol 2013; 32: 76-115. doi: 10.1097/
PGP.0b013e31826916c7. Erratum in: Int J Gynecol Pathol 2013; 32: 432. 
Erratum in: Int J Gynecol Pathol 2013; 32: 241.
7. Schiffman M, Doorbar J, Wentzensen N, de Sanjosé S, Fakhry C, Monk BJ, et 
al. Carcinogenic human papillomavirus infection. Nat Rev Dis Primers 2016; 
2: 16086. doi: 10.1038/nrdp.2016.86
8. Klaes R, Friedrich T, Spitkovsky D, Ridder R, Rudy W, Petry U, et al. 
Overexpression of p16(INK4A) as a specific marker for dysplastic and neo-
plastic epithelial cells of the cervix uteri. Int J Cancer 2001; 92: 276-84. doi: 
10.1002/ijc.1174
9. Tainio K, Athanasiou A, Tikkinen KAO, Aaltonen R, Cárdenas J, Hernándes, 
et al. Clinical course of untreated cervical intraepithelial neoplasia grade 2 
under active surveillance: systematic review and meta-analysis. BMJ 2018; 
360: k499. doi: 10.1136/bmj.k499
10. Mccredie MRE, Sharples KJ, Paul C, Baranyai J, Medley G, Jones RW, et al. 
Articles natural history of cervical neoplasia and risk of invasive cancer in 
women with cervical intraepithelial neoplasia 3: a retrospective cohort 
study. Lancet Oncol 2008; 9: 425-34. doi: 10.1016/S1470-2045(08)70103-7
11. Mills AM, Paquette C, Castle PE, Stoler MH. Risk stratification by p16 im-
munostaining of CIN1 biopsies: a retrospective study of patients from the 
quadrivalent HPV vaccine trials. Am J Surg Pathol 2015; 39: 611-7. doi: 
10.1097/PAS.0000000000000374
12. Sagasta A, Castillo P, Saco A, Torné A, Esteve R, Marimon L, et al. p16 stain-
ing has limited value in predicting the outcome of histological low-grade 
squamous intraepithelial lesions of the cervix. Mod Pathol 2016; 29: 51-9. 
doi: 10.1038/modpathol.2015.126
13. Barken SS, Rebolj M, Andersen ES, Lynge E. Frequency of cervical intraepi-
thelial neoplasia treatment in a well-screened population. Int J Cancer 2012; 
130: 2438-44. doi: 10.1002/ijc.26248
14. de Vicente JC, Rodríguez-Santamarta T, Rodrigo JP, Allonca E, Vallina A, 
Singhania A, et al. The emerging role of NANOG as an early cancer risk 
biomarker in patients with oral potentially malignant disorders. J Clin Med 
2019; 8: 1376. doi: 10.3390/jcm8091376
15. Rodrigo JP, Villaronga MÁ, Menéndez ST, Hermida-Prado F, Quer M, Vilaseca 
I, et al. A novel role for NANOG as an early cancer risk marker in patients 
with laryngeal precancerous lesions. Sci Rep 2017; 7: 11110. doi: 10.1038/
s41598-017-11709-9
16. Huang R, Rofstad EK. Cancer stem cells (CSCs), cervical CSCs and targeted 
therapies. Oncotarget 2017; 8: 35351-67. doi: 10.18632/oncotarget.10169
17. Takebe N, Percy Ivy S. Controversies in cancer stem cells: targeting em-
bryonic signaling pathways. Clin Cancer Res 2010; 16: 3106-12. doi: 
10.1158/1078-0432.CCR-09-2934
18. Prado MM, Frampton AE, Stebbing J, Krell J. Gene of the month: NANOG. J 
Clin Pathol 2015; 68: 763-5. doi: 10.1136/jclinpath-2015-203238
19. Maier S, Wilbertz T, Braun M, Scheble V, Reischl M, Mikut R, et al. SOX2 
amplification is a common event in squamous cell carcinomas of differ-
ent organ sites. Hum Pathol 2011; 42: 1078-88. doi: 10.1016/j.hump-
ath.2010.11.010
20. Porter L, McCaughan F. SOX2 and squamous cancers. Semin Cancer Biol 
2020. 67: 154-67. doi: 10.1016/j.semcancer.2020.05.007
21. Ji J, Wei X, Wang Y. Embryonic stem cell markers Sox-2 and OCT4 expression 
and their correlation with WNT signal pathway in cervical squamous cell 
carcinoma. Int J Clin Exp Pathol 2014; 7: 2470-6. PMID: 24966958
22. Xu R, Yang WT, Zheng PS. Coexpression of B-lymphoma Moloney murine 
leukemia virus insertion region-1 and sex-determining region of Y chromo-
some-related high mobility group box-2 in cervical carcinogenesis. Hum 
Pathol 2013; 44: 208-17. doi: 10.1016/j.humpath.2012.02.020
23. Ji J, Zheng PS. Expression of Sox2 in human cervical carcinogenesis. Hum 
Pathol 2010; 41: 1438-47. doi: 10.1016/j.humpath.2009.11.021
24. Yang Z, Pan X, Gao A, Zhu W. Expression of Sox2 in cervical squamous cell 
carcinoma. J BUON 2014; 19: 203-6. PMID: 24659665
25. Shen L, Huang X, Xie X, Su J, Yuan J, Chen X. High Expression of SOX2 and 
OCT4 Indicates radiation resistance and an independent negative prognosis 
in cervical squamous cell carcinoma. J Histochem Cytochem 2014; 62: 499-
509. doi: 10.1369/0022155414532654
26. Moshi JM, Hoogduin KJ, Ummelen M, Henfling MER, van Engeland M, 
Wouters KAD, et al. Switches of SOX17 and SOX2 expression in the devel-
opment of squamous metaplasia and squamous intraepithelial lesions of 
the uterine cervix. Cancer Med 2020; 9: 6330-43. doi: 10.1002/cam4.3201
Radiol Oncol 2025
Koren M et al. / Stem cell markers NANOG and SOX2 in the cervical squamous carcinogenesis12
27. Hou T, Zhang W, Tong C, Kazobinka G, Huang X, Huang Y, et al. Putative 
stem cell markers in cervical squamous cell carcinoma are correlated with 
poor clinical outcome. BMC Cancer 2015 24; 15: 785. doi: 10.1186/s12885-
015-1826-4
28. Kim BW, Cho H, Choi CH, Ylaya K, Chung JY, Kim JH, et al. Clinical significance 
of OCT4 and SOX2 protein expression in cervical cancer. BMC Cancer 2015; 
15: 1015. doi: 10.1186/s12885-015-2015-1
29. Wolsky RJ, Harbour LN, Mirza KM, Montag AG, Gwin K. The stem cell-
associated transcription factor SOX2 as a diagnostic marker of high-grade 
squamous intraepithelial lesion of the uterine cervix in comparison with 
p16 and Ki-67. Appl Immunohistochem Mol Morphol 2018; 26: 403-10. doi: 
10.1097/PAI.0000000000000434
30. Stewart CJ, Crook M. SOX2 expression in cervical intraepithelial neoplasia 
Grade 3 (CIN3) and superficially invasive (Stage IA1) squamous carci-
noma of the cervix. Int J Gynecol Pathol 2016; 35: 566-73. doi: 10.1097/
PGP.0000000000000273
31. Atıgan A, Kılıç D, Karakaya YA, Gök S, Güler ÖT. The relationship of immu-
nohistochemical SOX-2 staining with histopathological diagnosis in patients 
with abnormal colposcopic findings. Histochem Cell Biol 2023;1 60: 555-61. 
doi: 10.1007/s00418-023-02230-4
32. Moshi JM, Ummelen M, Broers JLV, Smedts F, van de Vijver KK, Cleutjens 
JPM, et al. SOX2 expression in the pathogenesis of premalignant lesions 
of the uterine cervix: its histo-topographical distribution distinguishes be-
tween low- and high-grade CIN. Histochem Cell Biol 2022; 158: 545-59. doi: 
10.1007/s00418-022-02145-6
33. Javed S, Sood S, Rai B, Bhattacharyya S, Bagga R, Srinivasan R. ALDH1 & 
CD133 in invasive cervical carcinoma & their association with the outcome 
of chemoradiation therapy. Indian J Med Res. 2021; 154: 367-74. doi: 
10.4103/ijmr.IJMR_709_20
34. Gu TT, Liu SY, Zheng PS. Cytoplasmic NANOG-positive stromal cells promote 
human cervical cancer progression. Am J Pathol 2012; 181: 652-61. doi: 
10.1016/j.ajpath.2012.04.008
35. Noh KH, Kim BW, Song KH, Cho H, Lee YH, Kim JH, et al. Nanog signaling in 
cancer promotes stem-like phenotype and immune evasion. J Clin Invest 
2012; 122: 4077-93. doi: 10.1172/JCI64057
36. Ye F, Zhou C, Cheng Q, Shen J, Chen H. Stem-cell-abundant proteins Nanog, 
Nucleostemin and Musashi1 are highly expressed in malignant cervical 
epithelial cells. BMC Cancer 2008; 8: 108. doi: 10.1186/1471-2407-8-108
37. Grubelnik G, Boštjančič E, Grošelj A, Zidar N. Expression of NANOG and its 
regulation in oral squamous cell carcinoma. Biomed Res Int 2020; 2020: 
8573793. doi: 10.1155/2020/8573793 
38. Lillsunde Larsson G, Carlsson J, Karlsson MG, Helenius G. Evaluation of HPV 
genotyping assays for archival clinical samples. J Mol Diagn 2015; 17: 293-
301. doi: 10.1016/j.jmoldx.2014.12.004
39. Seegene Inc. Anyplex™ II HPV28 detection user manual. Seoul: Seegene 
Inc; 2016. 
40. van Zummeren M, Leeman A, Kremer WW, Bleeker MCG, Jenkins D, van de 
Sandt M, et al. Three-tiered score for Ki-67 and p16ink4a improves accuracy 
and reproducibility of grading CIN lesions. J Clin Pathol 2018; 71: 981-8. doi: 
10.1136/jclinpath-2018-205271
41. Garland SM, Iftner T, Cuschieri K, Kaufmann AM, Arbyn M, de Sanjose S, 
et al; IPVS Policy Committee. IPVS policy statement on HPV nucleic acid 
testing guidance for those utilising/considering HPV as primary precancer 
screening: quality assurance and quality control issues. J Clin Virol 2023; 
159: 105349. doi: 10.1016/j.jcv.2022.105349
42. Vasefifar P, Motafakkerazad R, Maleki LA, Najafi S, Ghrobaninezhad F, 
Najafzadeh B, et al. Nanog, as a key cancer stem cell marker in tumor 
progression. Gene 2022; 827: 146448. doi: 10.1016/j.gene.2022.146448
43. Grubelnik G, Boštjančič E, Pavlič A, Kos M, Zidar N. NANOG expression in 
human development and cancerogenesis. Exp Biol Med (Maywood) 2020; 
245: 456-64. doi: 10.1177/1535370220905560
44. Bruni L, Albero G, Serrano B, Mena M, Collado JJ, Gómez D, et al. 
Human papillomavirus and related diseases in the world. Summary report 
[Internet]. Barcelona: ICO/IARC Information Centre on HPV and Cancer 
(HPV Information Centre); 2023. [cited 2025 Feb 28]. Available from: 
https://hpvcentre.net/statistics/reports/XWX.pdf
45. Chaturvedi AK, Katki HA, Hildesheim A, Rodríguez AC, Quint W, Schiffman 
M, et al; CVT Group. Human papillomavirus infection with multiple types: 
pattern of coinfection and risk of cervical disease. J Infect Dis 2011;203: 910-
20. doi: 10.1093/infdis/jiq139
46. Woodman CB, Collins SI, Young LS. The natural history of cervical HPV 
infection: unresolved issues. Nat Rev Cancer 2007; 7: 11-22. doi: 10.1038/
nrc2050
47. Organista-Nava J, Gómez-Gómez Y, Ocadiz-Delgado R, García-Villa E, Bonilla-
Delgado J, Lagunas-Martínez A, et al. The HPV16 E7 oncoprotein increases 
the expression of Oct3/4 and stemness-related genes and augments cell 
self-renewal. Virology 2016; 499: 230-42. doi: 10.1016/j.virol.2016.09.020
48. Kareta MS, Gorges LL, Hafeez S, Benayoun BA, Marro S, Zmoos AF, et al. 
Inhibition of pluripotency networks by the Rb tumor suppressor restricts 
reprogramming and tumorigenesis. Cell Stem Cell 2015; 16: 39-50. doi: 
10.1016/j.stem.2014.10.019
49. Lin T, Chao C, Saito S, Mazur SJ, Murphy ME, Appella E, et al. p53 induces 
differentiation of mouse embryonic stem cells by suppressing Nanog ex-
pression. Nat Cell Biol 2005; 7: 165-71. doi: 10.1038/ncb1211
50. Tyagi A, Vishnoi K, Mahata S, Verma G, Srivastava Y, Masaldan S, et al. 
Cervical cancer stem cells selectively overexpress HPV oncoprotein E6 that 
controls stemness and self-renewal through upregulation of HES1. Clin 
Cancer Res 2016; 22: 4170-84. doi: 10.1158/1078-0432.CCR-15-2574
51. Xi R, Pan S, Chen X, Hui B, Zhang L, Fu S, et al. HPV16 E6-E7 induces cancer 
stem-like cells phenotypes in esophageal squamous cell carcinoma through 
the activation of PI3K/Akt signaling pathway in vitro and in vivo. Oncotarget 
2016; 7: 57050-65. doi: 10.18632/oncotarget.10959
52. Martínez-Ramírez I, del-Castillo-Falconi V, Mitre-Aguilar IB, Amador-Molina 
A, Carrillo-García A, Langley E, et al. SOX2 as a new regulator of HPV16 
transcription. Viruses 2017; 9: 175. doi: 10.3390/v9070175
53. Díaz-Tejeda Y, Guido-Jiménez MC, López-Carbajal H, Amador-Molina A, 
Méndez-Martínez R, Gariglio-Vidal P, et al. Nanog, in cooperation with AP1, 
increases the expression of E6/E7 oncogenes from HPV types 16/18. Viruses 
2021; 13: 1482. doi: 10.3390/v13081482
54. Hart AH, Hartley L, Parker K, Ibrahim M, Looijenga LHJ, Pauchnik M, et al. 
The pluripotency homeobox gene NANOG is expressed in human germ cell 
tumors. Cancer 2005; 104: 2092-8. doi: 10.1002/cncr.21435
55. Chang DF, Tsai SC, Wang XC, Xia P, Senadheera D, Lutzko C. Molecular char-
acterization of the human NANOG protein. Stem Cells 2009; 27: 812-21. doi: 
10.1634/stemcells.2008-0657
56. Mikulenkova E, Neradil J, Vymazal O, Skoda J, Veselska R. NANOG/
NANOGP8 localizes at the centrosome and is spatiotemporally associated 
with centriole maturation. Cells 2020; 9: 692. doi: 10.3390/cells9030692
57. Driessens G, Blanpain C. Long live sox2: sox2 lasts a lifetime. Cell Stem Cell 
2011; 9: 283-4. doi: 10.1016/j.stem.2011.09.007
58. Mills AM, Coppock JD, Willis BC, Stoler MH. HPV E6/E7 mRNA in situ 
hybridization in the diagnosis of cervical low-grade squamous intraepi-
thelial lesions (LSIL). Am J Surg Pathol 2018; 42: 192-200. doi: 10.1097/
PAS.0000000000000974
59. Massad LS, Jeronimo J, Katki HA, Schiffman M; National Institutes of Health/
American Society for Colposcopy and Cervical Pathology Research Group. 
The accuracy of colposcopic grading for detection of high-grade cervical in-
traepithelial neoplasia. J Low Genit Tract Dis 2009; 13: 137-44. doi: 10.1097/
LGT.0b013e31819308d4