Access to this full-text is provided by MDPI.
Content available from Cancers
This content is subject to copyright.
Citation: Bandzeviˇci¯
ut˙
e, R.;
Platkeviˇcius, G.; ˇ
Ceponkus, J.; Želvys,
A.; ˇ
Cekauskas, A.; Šablinskas, V.
Differentiation of Urothelial
Carcinoma and Normal Bladder
Tissues by Means of Fiber-Based ATR
IR Spectroscopy. Cancers 2023,15, 499.
https://doi.org/10.3390/
cancers15020499
Academic Editors: Bartosz
Małkiewicz and Jakub Dobruch
Received: 26 October 2022
Revised: 6 January 2023
Accepted: 11 January 2023
Published: 13 January 2023
Copyright: © 2023 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
cancers
Article
Differentiation of Urothelial Carcinoma and Normal Bladder
Tissues by Means of Fiber-Based ATR IR Spectroscopy
Rimant˙
e Bandzeviˇci¯
ut˙
e1,* , Gediminas Platkeviˇcius 2,*, Justinas ˇ
Ceponkus 1, Ar¯
unas Želvys 2,
Albertas ˇ
Cekauskas 2and Valdas Šablinskas 1
1Institute of Chemical Physics, Faculty of Physics, Vilnius University, Saul˙
etekio av. 3,
LT-10257 Vilnius, Lithuania
2Institute of Clinical Medicine, Faculty of Medicine, Vilnius University, M. K. ˇ
Ciurlionio g. 21/27,
LT-03101 Vilnius, Lithuania
*Correspondence: rimante.bandzeviciute@ff.vu.lt (R.B.); gediminas.platkevicius@santa.lt (G.P.)
Simple Summary:
Non-muscle-invasive bladder cancer (NMIBC) is a complicated locally treated
disease with high recurrence rates and high risk of progression to muscle invasive disease. Current
standard diagnostic approach based on white light cystoscopy (WLC) is characterized by high false
negative rate, excessive invasion of the bladder and high economic burden. Hereby we present the
fiber-based attenuated total reflection infrared (ATR IR) spectroscopy study of healthy and cancerous
tissue samples taken from 54 patients to delineate normal and tumorous human bladder tissues
under ex vivo conditions. Investigation of the tissue samples immediately after surgical operation
allows to examine samples in their native conditions and establish their initial structure and chemical
composition avoiding sample degradation. Our study reveals that fiber-based ATR IR spectroscopy
could be an addition to current endoscopic approach with increased diagnostic accuracy and with
some potential to be applied in vivo.
Abstract:
Surgical treatment is widely applied curative approach for bladder cancer. White light cys-
toscopy (WLC) is currently used for intraoperative diagnostics of malignant lesions but has relatively
high false-negative rate. Here we represent an application of label free fiber-based attenuated total
reflection infrared spectroscopy (ATR IR) for freshly resected human bladder tissue examination for
54 patients. Defined molecular spectral markers allow to identify normal and urothelial carcinoma
tissues. While methods of statistical analysis (Hierarchical cluster analysis (HCA) and Principal
component analysis (PCA)) used for spectral data treatment allow to discriminate tissue types with
91% sensitivity and 96–98% specificity. In the present study the described method was applied for
tissue examination under ex vivo conditions. However, after method validation the equipment could
be translated from laboratory studies to in situ or even in vivo studies in operating room.
Keywords:
bladder cancer; urothelial carcinoma; fiber-based ATR IR spectroscopy; molecular markers
1. Introduction
Bladder cancer (BC) is the tenth most diagnosed cancer worldwide, with approxi-
mately 573,000 new cases and 213,000 deaths annually. It is approximately four times
more common in men than in women, with respective incidence and mortality rates of
9.5 and 3.3 per 100,000 among men [
1
]. Non-muscle-invasive bladder cancer (NMIBC)
locally treated and highly recurrent disease which involves only the urothelium or lamina
propria without invasion to the detrusor muscle, accounts for about 80% of all bladder
cancers [2].
The 5-year recurrence and progression rates depend on clinical and pathological
factors and varies from 31% to 78% and from 0.8% to 45%, respectively [
3
]. Because of
the high recurrence rate and complexity of the invasive diagnostic procedures, NMIBC is
Cancers 2023,15, 499. https://doi.org/10.3390/cancers15020499 https://www.mdpi.com/journal/cancers
Cancers 2023,15, 499 2 of 11
associated with a high humanistic and economic burden to maintain per patient over his
lifetime [4].
White light cystoscopy (WLC), endoscopy of the urinary bladder via urethra, combined
with transurethral resection of the bladder (TURB) procedure and subsequent pathology
examination of the specimens is the mainstay of diagnosis in BC. This diagnostic approach
enables to access the local staging and grading of the disease to determine the optimal patient
management tactic [
5
]. However, WLC has a negative predictive value (NPV) of 56% [
6
],
which could be increased up to 86% in combination with urinary cytology examination [7].
Furthermore, it requires surgical resection of the bladder wall. Complication rates
are approximately 4–6% of which urinary tract infections and significant hematuria are
the most common [
8
]. Some of these resections are not necessary, as we know from the
literature that positive predictive value (PPV) of WLC alone is only 64% [
6
]. As a result,
novel technologies are under development to improve lesion detection, diagnostic accuracy,
and prognosis, when having minimal risk of complications and reduced economic burden.
One such marker free alternative is a narrow band imaging cystoscopy (NBI). NBI
uses light at different wavelengths than WLC and allow to see hyper vascularized areas of
the mucosa more clearly in contrast with healthy tissue. Studies that have compared NBI
to WLC show, that NBI increases sensitivity, however, has a significantly lower specificity
(50% vs. 86.9%, p< 0.001) and an increase in unnecessary biopsies [9].
A variety of optical spectroscopy methods including fluorescence, Raman scattering,
and infrared (IR) absorption spectroscopy becomes more and more widely applied for the
investigation of biological tissues in the field of surgical oncology. Methods of fluorescence
are already applied for the surgical procedures. Fluorescence guided surgery is already
being used for several types of oncological surgery [
10
]. During the surgery, this method
allows to visualize different tissue types such as normal tissue and malignant lesions that
would not be visible under the naked eye inspection. However, there are some limitations.
In order to visualize different tissue types, fluorescence contrast agent should be inserted
into the body before the investigation to be accumulated in the targeted tumorous tissue
area. In this regard, vibrational spectroscopy methods (Raman scattering and IR absorption)
have some advantages since they can be considered as a label-free methods without
requirement of tissue dyeing or chemical processing prior the examination.
Methods of spontaneous and coherent Raman scattering spectroscopy were applied
for the tumorous tissue identification in human and animal models and showed promising
results [
11
]. In particular, applied fiber-based technique for the brain tissue examination
showed the ability to perform the investigation under
in vivo
conditions and showed
good results for the intraoperative tumor detection [
12
]. However, Raman spectroscopy
has some limitations. Raman scattering is a relatively weak process therefore the signal
of Raman scattering usually is not strong and particularly sensitive detectors should be
used in order to receive a sufficient Raman signal. Moreover, depending on the wavelength
and power of the laser that is used for the excitation of the Raman scattering, undesirable
effects of fluorescence background in the Raman spectra can be observed. Attention to
the surrounding light sources also should be paid: Radiation from the other light sources
used in the operating room could interfere with the operating mode of the detector and
have influence to the obtained spectra. Additionally, the usage of laser radiation in Raman
spectroscopy can be problematic: the power of the laser and the integration time of the
measurement should be carefully selected in order to avoid the damage of the tissue, and
at the same time obtain the sufficient Raman signal and to receive the data in a reasonable
time to maintain the examination time as short as possible without prolonging the surgery
and anesthesia time.
Meanwhile, IR radiation is non-destructive for biological tissues and a spectrum can be
obtained quite fast. IR spectroscopy is already used for studies of various biological samples
including human tumorous tissue or cells [
13
–
19
]. Applied methods cover conventional
techniques such as IR transmission, micro spectroscopy and attenuated total reflection
(ATR). In most cases, conventional methods require a special sample preparation as well
Cancers 2023,15, 499 3 of 11
as placement of the samples to the special sample compartment inside the spectrometer.
In addition, standard spectrometers usually are bulky and non-maneuverable that is an
obstacle for the implementation of the method into the operating room. Application of
fiber optics for the IR spectroscopy enables to bring the method one step forward clinical
diagnostics. This provides the capability of performing the measurements of spectra under
in situ conditions without the requirement to transfer the tissue into the device. Moreover,
mobile devices of small dimensions could be coupled with fiber optics thus allowing to
transfer the equipment almost anywhere in the operating room without the disturbing the
flow of the surgery.
Some authors report results of application of IR spectroscopy for human BC investi-
gation, however, most of them cover applications for specially prepared tissue samples
such as formalin-fixed and paraffin-embedded tissues [
20
,
21
] or cultured cell lines fixed
in glutaraldehyde in Phosphate-Buffered Saline (PBS) [
22
]. In our previous studies we
showed that fiber-based IR spectroscopy can be successfully applied for examination of
various freshly resected human tissues [
23
,
24
]. Here we represent the study of 54 patients
by using fiber-based IR spectroscopy to delineate normal and tumorous human bladder
tissues under ex vivo conditions. To our knowledge, in this manuscript we for the first time
present the study of fiber probe application for freshly resected untreated human bladder
tissue samples. Investigation of the tissue samples immediately after surgical operation
allows to examine samples in their native conditions and establish their initial structure
and chemical composition avoiding sample degradation. After further implementations,
the method could be used even
in vivo
conditions thus allowing easier tissue examination
during the surgery.
2. Materials and Methods
2.1. Sample Collection
Spectral studies of the bladder tissues were approved by the Regional Biomedical
Research Ethics Committee (Document No. 2019/12-1178-665). The details of the sample
collection methodic is described in our previous paper [
25
]. Briefly, the samples of the
bladder tissues for fiber-based ATR IR spectroscopic studies were obtained between July
2019 and September 2021 in the tertiary Urology Center when performing TURB procedure.
Patients were eligible if they had a clinical or radiological suspicion of BC, and they were
required TURB procedure according to the latest NMIBC guidelines of the European
Association of Urology.
Samples were collected during the TURB procedure in the following manner: we
began by obtaining single sample of healthy-looking bladder tissue, then–single sample of
the bladder superficial tumor. After the procedure, bladder tissue samples were sent for
histological and spectroscopic analysis. Tumor malignancy was confirmed by a pathologist
by examining the resected tissue.
2.2. Spectroscopy
IR spectra of freshly resected tissue samples were collected using ATR IR fiber probe.
The system consists of the ATR silver halide fiber probe (Art Photonics GmbH, Berlin,
Germany) attached to the standard FT-IR (Fourier transform infrared) spectrometer Alpha
(Bruker Optik GmbH, Ettlingen, Germany) additionally equipped with the external liquid
nitrogen cooled MCT (Mercury-Cadmium-Telluride) detector (Infrared Associates, Inc.
Model IRA-20-00131).
Small cuts of normal and tumorous bladder tissue were made. Spectra collected
by pressing the ATR fiber probe to the surface of freshly cut tissue. For each sample
several spectra were measured and were averaged for the analysis. Spectra collected in
the spectral region between 400 and 4000 cm
−1
with 4 cm
−1
spectral resolution. Sixty-
four interferograms were averaged and Fourier transformed into a spectrum applying
Blackmann-Harris 3 term apodization function and zero filling factor of 2. Before every
Cancers 2023,15, 499 4 of 11
measurement ATR crystal was cleaned with distilled water and ethanol and the background
spectrum of the ambient air was measured.
2.3. Statistical Analysis
Statistical analysis (Hierarchical cluster analysis (HCA) and Principal component
analysis (PCA)) was performed by using data analysis software OriginPro (OriginLab cor-
poration). Prior statistical analysis, spectra were pre-processed: atmospheric compensation,
baseline correction, vector normalization and offset correction were applied by using spec-
troscopy software OPUS (Bruker Optik GmbH, Ettlingen, Germany). Before performing
HCA analysis, first derivation of spectra was calculated by using spectroscopy software
OPUS. Statistical analysis was performed by applying standard procedures (Hierarchical
Cluster Analysis and Principal Component Analysis) of the OriginPro software package.
While performing HCA, Ward cluster algorithm, Euclidean distance type were chosen.
For both statistical methods (HCA and PCA), analysis was performed in 950–1480 cm
−1
spectral range. Different spectral regions were chosen for statistical analysis; however, the
best results were achieved by performing analysis in the mentioned spectral range.
In order to evaluate the reliability of the method, sensitivity, specificity, positive
predictive value (PPV) and negative predictive value (NPV) were calculated as follows:
Sensitivity =True positives
Truepositives +Falsenegatives,
Specificity =True negatives
Truenegatives +Falsepositives,
PPV =True positives
Truepositives +Falsepositives,
NPV =True negatives
Truenegatives +Falsenegatives.
95% Confidence intervals (CI) were calculated by using MedCalc (MedCalc Software
Ltd., Ostend, Belgium) online tool [26] for statistical analysis.
In total, 54 patient cases were analyzed. The summarized data is shown in the Table 1.
In most cases, two types of tissue samples (normal and tumorous) were received for spectral
analysis per patient. In several cases, only normal or only tumorous tissue was available for
analysis from particular patient, thus only 49 normal and 47 tumorous tissue samples were
analyzed. More detailed information is presented in Supplementary material (Table S1).
Table 1.
Tissue types collected and patient characteristics during the study. Information based on
histological analysis, surgery information.
Tissue Type Number of Samples
Normal tissue 49
Tumorous tissue
High-grade tumor 34
Low-grade tumor 13
Patient characteristics Number of patients
Tumor size during initial TURB
<10 mm 12
10–30 mm 18
>30 mm 17
Tumor count on initial TURB
Single 15
Two to three 15
More than three 17
Cancers 2023,15, 499 5 of 11
Table 1. Cont.
Tissue Type Number of Samples
Additional information
Number of patients
Median patient age (years)
54
70.5
3. Results
Mean ATR IR spectra of normal and tumorous bladder tissues presented in Figure 1
demonstrate obviously different profiles. Main spectral band positions and their assign-
ments are shown in Table 2.
Cancers 2023, 15, x FOR PEER REVIEW 5 of 11
10–30 mm 18
>30 mm 17
Tumor count on initial TURB
Single 15
Two to three 15
More than three 17
Additional information
Number of patients
Median patient age (years)
54
70.5
3. Results
Mean ATR IR spectra of normal and tumorous bladder tissues presented in Figure 1
demonstrate obviously different profiles. Main spectral band positions and their assign-
ments are shown in Table 2.
Figure 1. Mean ATR IR spectra of normal and urothelial carcinoma (high-grade tumor and low-
grade tumor) tissue. Spectra are baseline corrected and vector normalized.
Table 2. Spectral bands and their assignments [27].
Spectral Band Position, cm−1 Assignment
Normal Tissue Urothelial Carcinoma Tissue
972 972
ν(PO4) of nucleic acids and proteins
OCH3 of polysaccharides
Not present 1028 ν(C-O), ν(C-C), δ(C-O-H) of glycogen
1033 Not present Collagen
1050 1048 (low-grade)
1052 (high-grade)
ν(CO-O-C)
ν(C-O), δ(C-O) of the C-OH of carbo-
hydrates
1082 1082 ν(PO−2) of nucleic acids
Shoulder at 1121 Shoulder at 1121
ν
(C-O), δ(C-O), δ(C-O-H), δ(C-O-C)
phosphodiester stretching
Figure 1.
Mean ATR IR spectra of normal and urothelial carcinoma (high-grade tumor and low-grade
tumor) tissue. Spectra are baseline corrected and vector normalized.
Mean spectrum of normal bladder tissue represents more prominent spectral bands lo-
cated at 1033, 1206, 1240, 1282, 1317 and 1339 cm
−1
which can be assigned to collagen
[27,28]
and indicates higher collagen levels in normal bladder tissue. Spectra of tumorous bladder
tissue demonstrate higher absorbance values of the spectral band located at 972 cm
−1
which could be assigned to
ν
(PO
4
) vibrations of nucleic acids. Moreover, spectral bands
located at similar positions can be assigned to
ν
(PO
4
) vibrations of proteins and vibrational
modes of OCH
3
groups of polysaccharides. However, herewith the increased absorbance
values of this spectral band, the stronger absorption is for the spectral band located at
1082 cm
−1
in tumorous tissue spectra and assigned to
ν
(PO
−
2
) vibrations of nucleic acids.
These findings can be related to increased levels of nucleic acids in urothelial carcinoma
tissue. Higher values of absorbance for the band positions at 1028 cm
−1
assigned to
ν
(C-O),
ν
(C-C),
δ
(C-O-H) vibrations and at 1154 cm
−1
assigned to
ν
(C-O) vibrations of glycogen
are observed.
Although differences between the mean spectra of normal and tumorous tissues are
clearly visible and well defined, variations of spectra inside both tissue classes (normal and
tumorous tissues) are quite high (Figure 2).
Cancers 2023,15, 499 6 of 11
Table 2. Spectral bands and their assignments [27].
Spectral Band Position, cm−1
Assignment
Normal Tissue Urothelial Carcinoma Tissue
972 972 ν(PO4) of nucleic acids and proteins
OCH3of polysaccharides
Not present 1028 ν(C-O), ν(C-C), δ(C-O-H) of glycogen
1033 Not present Collagen
1050 1048 (low-grade)
1052 (high-grade)
ν(CO-O-C)
ν(C-O), δ(C-O) of the C-OH of carbohydrates
1082 1082 ν(PO−
2) of nucleic acids
Shoulder at 1121 Shoulder at 1121 ν(C-O), δ(C-O), δ(C-O-H), δ(C-O-C)
phosphodiester stretching
Shoulder at 1154 1154 ν(C-O)
1163 Not present ν(C-O)
Shoulder at 1172 1172 ν(C-O)
1206 Not present Amide III
1241 1239 ν(PO−
2), Amide III
1282 1284 Amide III
Not present 1307 ν(CH2)
1317 Shoulder at 1317 Amide III
1339 1339 CH2wagging
1399 1399 δ(CH3)
1457 1457 δ(CH3)
Cancers 2023, 15, x FOR PEER REVIEW 6 of 11
Shoulder at 1154 1154 ν(C-O)
1163 Not present ν(C-O)
Shoulder at 1172 1172
ν
(C-O)
1206 Not present Amide III
1241 1239 ν(PO−2), Amide III
1282 1284 Amide III
Not present 1307 ν(CH2)
1317 Shoulder at 1317 Amide III
1339 1339 CH2 wagging
1399 1399 δ(CH3)
1457 1457 δ(CH3)
Mean spectrum of normal bladder tissue represents more prominent spectral bands
located at 1033, 1206, 1240, 1282, 1317 and 1339 cm−1 which can be assigned to collagen
[27,28] and indicates higher collagen levels in normal bladder tissue. Spectra of tumorous
bladder tissue demonstrate higher absorbance values of the spectral band located at 972
cm−1 which could be assigned to ν(PO4) vibrations of nucleic acids. Moreover, spectral
bands located at similar positions can be assigned to ν(PO4) vibrations of proteins and
vibrational modes of OCH3 groups of polysaccharides. However, herewith the increased
absorbance values of this spectral band, the stronger absorption is for the spectral band
located at 1082 cm−1 in tumorous tissue spectra and assigned to ν(PO−2) vibrations of nu-
cleic acids. These findings can be related to increased levels of nucleic acids in urothelial
carcinoma tissue. Higher values of absorbance for the band positions at 1028 cm−1 assigned
to ν(C-O), ν(C-C), δ(C-O-H) vibrations and at 1154 cm−1 assigned to ν(C-O) vibrations of
glycogen are observed.
Although differences between the mean spectra of normal and tumorous tissues are
clearly visible and well defined, variations of spectra inside both tissue classes (normal
and tumorous tissues) are quite high (Figure 2).
Figure 2. Mean spectra (black solid line) and variations of spectra of tissue taken from different
patients (grey area) of normal (a) and tumorous urothelial carcinoma (b) tissue.
Black lines correspond to the mean spectra of all the samples taken from different
patients, while the grey areas indicate the variations of spectra in different tissue classes.
Variations of spectra in the tumorous tissue class are higher (Figure 2b), especially for the
spectral bands corresponding to glycogen (located at 1028 and 1154 cm−1) and collagen
Figure 2.
Mean spectra (black solid line) and variations of spectra of tissue taken from different
patients (grey area) of normal (a) and tumorous urothelial carcinoma (b) tissue.
Black lines correspond to the mean spectra of all the samples taken from different
patients, while the grey areas indicate the variations of spectra in different tissue classes.
Variations of spectra in the tumorous tissue class are higher (Figure 2b), especially for the
spectral bands corresponding to glycogen (located at 1028 and 1154 cm
−1
) and collagen
(1240, 1284, 1317, 1339 and 1457 cm
−1
) and could be considered as spectral tumor markers
of urothelial carcinoma.
In order to evaluate the capability of the IR spectroscopy to delineate tumorous and
normal tissues, HCA and PCA analysis were applied. Figure 3shows the dendrogram
of HCA analysis. Spectra of tumorous and normal tissues are classified into 2 clusters
(red cluster on the left and blue cluster on the right side of the dendrogram correspond to
tumorous and normal tissue classes, respectively), large distance between clusters reveals
Cancers 2023,15, 499 7 of 11
a good separation between both tissue classes. Only 1 spectrum out of 49 normal tissue
spectra and 4 spectra out of 47 tumorous tissue spectra were assigned to wrong classes
(these spectra are indicated by * symbol) that corresponds to 91% sensitivity and 98%
specificity (Table 3).
Cancers 2023, 15, x FOR PEER REVIEW 7 of 11
(1240, 1284, 1317, 1339 and 1457 cm−1) and could be considered as spectral tumor markers
of urothelial carcinoma.
In order to evaluate the capability of the IR spectroscopy to delineate tumorous and
normal tissues, HCA and PCA analysis were applied. Figure 3 shows the dendrogram of
HCA analysis. Spectra of tumorous and normal tissues are classified into 2 clusters (red
cluster on the left and blue cluster on the right side of the dendrogram correspond to tu-
morous and normal tissue classes, respectively), large distance between clusters reveals a
good separation between both tissue classes. Only 1 spectrum out of 49 normal tissue
spectra and 4 spectra out of 47 tumorous tissue spectra were assigned to wrong classes
(these spectra are indicated by * symbol) that corresponds to 91% sensitivity and 98%
specificity (Table 3).
Figure 3. HCA dendrogram. Red cluster corresponds to the tumorous tissue class, blue cluster cor-
responds to normal tissue class. Letters T and N represent spectra of each tumorous and normal
tissue samples, respectively. * Symbol indicates spectra which were assigned to wrong tissue class.
Results of PCA analysis are presented in Figure 4. Spectra are normalized thus the
most part of variance is explained by first and second principal components (PCs) (per-
centage of variance is 39% and 16% for the first and second PCs, respectively). The most
significant chemical changes in the tissue samples are defined by first and second PCs,
thus the score plot is shown for these two PCs. Clusters on the left and right sides of the
diagram correspond to normal and tumorous tissue classes, respectively. Blue dots indi-
cate spectra of each normal tissue, red and orange dots indicate spectra of each tumorous
(high- and low-grade of malignancy, respectively) tissue samples. Classes are well sepa-
rated, only 2 spectra out of 49 normal tissue spectra and 4 spectra out of 47 tumorous
tissue spectra were assigned to wrong classes that corresponds to 91% sensitivity and 96%
specificity (Table 3).
Figure 3.
HCA dendrogram. Red cluster corresponds to the tumorous tissue class, blue cluster
corresponds to normal tissue class. Letters T and N represent spectra of each tumorous and normal
tissue samples, respectively. * Symbol indicates spectra which were assigned to wrong tissue class.
Table 3.
Summarized values of sensitivity, specificity, PPV and NPV values for HCA and PCA results.
Method
HCA PCA
Value 95% CI Value 95% CI
Sensitivity 91% 80% to 98% 91% 80% to 98%
Specificity 98% 89% to 100% 96% 86% to 100%
PPV 98% 86% to 100% 96% 85% to 99%
NPV 92% 82% to 97% 92% 87% to 98%
CI–confidence interval.
Results of PCA analysis are presented in Figure 4. Spectra are normalized thus the most
part of variance is explained by first and second principal components (PCs) (percentage of
variance is 39% and 16% for the first and second PCs, respectively). The most significant
chemical changes in the tissue samples are defined by first and second PCs, thus the
score plot is shown for these two PCs. Clusters on the left and right sides of the diagram
correspond to normal and tumorous tissue classes, respectively. Blue dots indicate spectra
of each normal tissue, red and orange dots indicate spectra of each tumorous (high- and
low-grade of malignancy, respectively) tissue samples. Classes are well separated, only
2 spectra out of 49 normal tissue spectra and 4 spectra out of 47 tumorous tissue spectra were
assigned to wrong classes that corresponds to 91% sensitivity and 96% specificity (Table 3).
Cancers 2023,15, 499 8 of 11
Cancers 2023, 15, x FOR PEER REVIEW 8 of 11
Figure 4. PCA diagram. Blue dots indicate spectra of each normal tissue samples, red and orange
dots indicate spectra of each tumorous (high- and low-grade of malignancy, respectively) tissue
samples. Percentage of variance for PCs is shown in brackets.
Table 3. Summarized values of sensitivity, specificity, PPV and NPV values for HCA and PCA re-
sults.
Method
HCA PCA
Value 95% CI Value 95% CI
Sensitivity 91% 80% to 98% 91% 80% to 98%
Specificity 98% 89% to 100% 96% 86% to 100%
PPV 98% 86% to 100% 96% 85% to 99%
NPV 92% 82% to 97% 92% 87% to 98%
CI–confidence interval.
4. Discussion
As it is observed fro m the ATR IR spectra (Figure 1), normal bladder tissues are richer
in collagen compared to urothelial carcinoma tissues. Collagen is one of the components
of the lamina propria of urinary bladder and assures the tensile strength by transferring
the tension from the bladder smooth muscle cells, during the expansion of the organ while
storing urine [29]. Reduced collagen levels in tumorous bladder tissue can be related to
increased enzymes secretion of collagenase into the tissue matrix and destruction of host
tissue [30]. Possibly higher levels of nucleic acids in urothelial carcinoma tissue can be
related to higher rates of cell proliferation due to growth of the tumor. Increased absorb-
ance values of spectral bands assigned to glycogen are related to increased amounts of
glycogen in tumorous bladder tissues. Increased levels of glycogen are observed in some
types of tumors and used as the resource of energy for the cell proliferation [31,32]. More-
over, higher levels of glycogen are observed in low-grade urothelial carcinoma tissue com-
pared to high-grade tumor tissue. Observed inverse relation of glycogen amount and tu-
mor grade could be related to higher proliferation rates of high-grade tumors. In some
types of tumors inverse relation of glycogen amount and proliferation rates are observed
and possibly linked to higher glycogen consumption for maintaining tumor growth [32].
Higher variations in tumorous tissue spectra class compared to normal tissue spectra
class could be linked to the diverse differentiation of tumorous cells. During development
Figure 4.
PCA diagram. Blue dots indicate spectra of each normal tissue samples, red and orange
dots indicate spectra of each tumorous (high- and low-grade of malignancy, respectively) tissue
samples. Percentage of variance for PCs is shown in brackets.
4. Discussion
As it is observed from the ATR IR spectra (Figure 1), normal bladder tissues are richer
in collagen compared to urothelial carcinoma tissues. Collagen is one of the components
of the lamina propria of urinary bladder and assures the tensile strength by transferring
the tension from the bladder smooth muscle cells, during the expansion of the organ while
storing urine [
29
]. Reduced collagen levels in tumorous bladder tissue can be related to
increased enzymes secretion of collagenase into the tissue matrix and destruction of host
tissue [
30
]. Possibly higher levels of nucleic acids in urothelial carcinoma tissue can be
related to higher rates of cell proliferation due to growth of the tumor. Increased absorbance
values of spectral bands assigned to glycogen are related to increased amounts of glycogen
in tumorous bladder tissues. Increased levels of glycogen are observed in some types of
tumors and used as the resource of energy for the cell proliferation [
31
,
32
]. Moreover,
higher levels of glycogen are observed in low-grade urothelial carcinoma tissue compared
to high-grade tumor tissue. Observed inverse relation of glycogen amount and tumor grade
could be related to higher proliferation rates of high-grade tumors. In some types of tumors
inverse relation of glycogen amount and proliferation rates are observed and possibly
linked to higher glycogen consumption for maintaining tumor growth [32].
Higher variations in tumorous tissue spectra class compared to normal tissue spectra
class could be linked to the diverse differentiation of tumorous cells. During development
of the tumor, cells lose their functions and change their morphological features. Thus,
depending on the tumor features and development, variations of the amounts of specific
components can vary in individual cases. Meanwhile variations in normal tissue spectra
class are smaller (Figure 2a) compared to tumorous tissue class. Spectral differences may
be caused by individual differences in tissue constitution.
In this study, presented sensitivity values reach 91% (95% CI 80–98%) (by applying
both HCA and PCA methods) while specificity values reach 98% (95% CI 89–100%) and
96% (95% CI 86–100%) by applying HCA and PCA methods, respectively. Currently,
most widely used endoscopic approach to diagnose BC is WLC with sensitivity of 71%
(95% CI, 49–93%) and specificity of 72% (95% CI, 47–96%) [
33
]. A novel technique with
instillation of photosensitizing agent, so called photodynamic diagnosis (PDD), has shown
better sensitivity rates of (92% (95% CI, 80–100%)) but lower specificity (57% (95% CI,
36–79%)) [
33
]. Another novel technique known as NBI is based on less invasive process of
Cancers 2023,15, 499 9 of 11
bladder wall illumination with filtered white light. It enhances the visualization of mucosal
blood vessels because emitted wavelengths are absorbed by hemoglobin more intensively
compared to other mucosal tissue and helps to contrast the neoangiogenic urothelial
tumors [
34
]. According to the literature, NBI offers high sensitivity (96% (
95% CI = 93–98%
))
but low specificity (65% (95% CI = 54–75%)), similarly to PDD [5].
Our study has showed promising results to differentiate healthy and cancerous tissue
of urinary bladder under ex vivo conditions, which are close to the golden standard of
pathological examination. Fiber-based ATR IR spectroscopy showed similar sensitivity
results as the results in the literature of PDD and NBI without a decrease in specificity. This
could lead to less invasive diagnostic approach, when correct diagnosis would not require
excessive resection of the bladder. Moreover, high specificity of the test could lead to a
decrease in numbers of pathological analyses required in the follow-up of the patient with
reduced economic burden.
However, our study shows some limitations. Firstly, we were unable to obtain both
cancerous and normal bladder tissue samples from every patient. Surgeries were performed
for patients with suspected malignant bladder neoplasms. Resected tissue was evaluated
macroscopically by a surgeon; subsequently, suspicious looking tissue was assigned as
tumorous and a sample of resected tissue was given for spectral analysis, the whole resected
tissue was sent for routine pathological examination. After pathological examination,
6 samples suspected as tumorous finally were identified as chronic cystitis. Dendrogram
of HCA analysis including these 6 cystitis cases is presented in supplementary material
(Figure S1). In one case it was impossible to obtain tumorous tissue. These 7 patient
cases were not included in the statistical analysis presented in the main text of this paper.
Other patients were diagnosed with completely altered bladder and we were unable to
obtain normal bladder tissue samples. In this case, 5 patients were improper to obtain
normal bladder tissue. Secondly, despite observed differences in mean spectra of low-grade
and high-grade tumors, spectral profiles of low- and high-grade cancerous tissue share
similar features thus not allowing to discriminate them into separate groups. This could
be explained by different number of low- and high-grade tumor cases included in the
study. Number of high-grade tumor cases is much bigger than low-grade tumor cases
(34 and 13 cases, respectively), thus for the more detailed conclusions a larger scope of
low-grade tumor cases would be beneficial. However, observed spectral features and
applied statistical analysis allow to discriminate urothelial carcinoma and normal bladder
tissues with high accuracy rates under ex vivo conditions on cystoscopically selected tissue
samples. Based on our study data, fiber-based ATR IR spectroscopy could be an addition
to current endoscopic approach with an increase in diagnostic accuracy, if successfully
applied in vivo.
5. Conclusions
This study shows the possible ex vivo application of the fiber-based ATR IR spec-
troscopy for the discrimination between the cancerous and normal human bladder tissues
with promising results for further studies. High rates of tissue identification accuracy
suggest that the applied method after some implementations could be used in addition to
current diagnostic approaches.
Supplementary Materials:
The following supporting information can be downloaded at: https:
//www.mdpi.com/article/10.3390/cancers15020499/s1. Table S1: Detailed patient and tissue sample
information. Figure S1: HCA dendrogram of benign and malignant tissues.
Author Contributions:
Conceptualization: R.B., G.P., J.ˇ
C., A.Ž., A. ˇ
C. and V.Š.; methodology: R.B.,
J. ˇ
C. and V.Š.; validation: R.B. and G.P.; formal analysis, R.B.; investigation, R.B. and G.P.; resources:
G.P., A.Ž. and A. ˇ
C.; data curation: R.B. and G.P.; writing—original draft preparation: R.B. and G.P.;
writing—review and editing: J. ˇ
C., A.Ž., A. ˇ
C. and V.Š.; visualization: R.B. All authors have read and
agreed to the published version of the manuscript.
Funding: This research received no external funding.
Cancers 2023,15, 499 10 of 11
Institutional Review Board Statement:
The study was approved by Ethics Committee of Vilnius
Regional Biomedical Research (original name: Vilniaus Regioninis Biomedicinini
u˛
Tyrim
u˛
Etikos
Komitetas) (Document No. 2019/12-1178-665 and date of approval: 3rd December 2019).
Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.
Data Availability Statement: The data can be shared up on request.
Acknowledgments:
We would like to thank all the study participants, who accepted to be included
in our study. Furthermore, we would like to thank all the urologists of Vilnius University Santaros
Clinics for their cooperation during the study.
Conflicts of Interest: The authors declare no conflict of interest.
References
1.
Sung, H.; Ferlay, J.; Siegel, R.L.; Laversanne, M.; Soerjomataram, I.; Jemal, A.; Bray, F. Global Cancer Statistics 2020: GLOBOCAN
Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J. Clin.
2021
,71, 209–249. [CrossRef]
[PubMed]
2.
Miyake, M.; Gotoh, D.; Shimada, K.; Tatsumi, Y.; Nakai, Y.; Anai, S.; Torimoto, K.; Aoki, K.; Tanaka, N.; Konishi, N.; et al.
Exploration of risk factors predicting outcomes for primary T1 high-grade bladder cancer and validation of the Spanish Urological
Club for Oncological Treatment scoring model: Long-term follow-up experience at a single institute. Int. J. Urol.
2015
,22, 541–547.
[CrossRef] [PubMed]
3.
Sylvester, R.J.; van der Meijden, A.P.; Oosterlinck, W.; Witjes, J.A.; Bouffioux, C.; Denis, L.; Newling, D.W.; Kurth, K. Predicting
Recurrence and Progression in Individual Patients with Stage Ta T1 Bladder Cancer Using EORTC Risk Tables: A Combined
Analysis of 2596 Patients from Seven EORTC Trials. Eur. Urol. 2006,49, 466–477. [CrossRef] [PubMed]
4.
Lee, L.J.; Kwon, C.S.; Forsythe, A.; Mamolo, C.M.; Masters, E.T.; Jacobs, I.A. Humanistic and Economic Burden of Non-Muscle
Invasive Bladder Cancer: Results of Two Systematic Literature Reviews. Clinicoecon. Outcomes Res.
2020
,12, 693–709. [CrossRef]
[PubMed]
5.
Russo, G.I.; Sholklapper, T.N.; Cocci, A.; Broggi, G.; Caltabiano, R.; Smith, A.B.; Lotan, Y.; Morgia, G.; Kamat, A.M.;
Witjes, J.A.; et al.
Performance of Narrow Band Imaging (NBI) and Photodynamic Diagnosis (PDD) Fluorescence Imaging
Compared to White Light Cystoscopy (WLC) in Detecting Non-Muscle Invasive Bladder Cancer: A Systematic Review and
Lesion-Level Diagnostic Meta-Analysis. Cancers 2021,13, 4378. [CrossRef]
6.
Daneshmand, S.; Bazargani, S.T.; Bivalacqua, T.T.; Holzbeierlein, J.M.; Willard, B.; Taylor J., M.; Liao, J.C.; Pohar, K.;
Tierney, J.
;
Konety, B. Blue light cystoscopy for the diagnosis of bladder cancer: Results from the US prospective multicenter registry.
Urol. Oncol.: Semin. Orig. Investig. 2018,36, 361.e1. [CrossRef]
7.
Grossman, H.B.; Soloway, M.; Messing, E.; Katz, G.; Stein, B.; Kassabian, V.; Shen, Y. Surveillance for recurrent bladder cancer
using a point-of-care proteomic assay. JAMA 2006,295, 299–305. [CrossRef]
8. Kim, L.H.C.; Patel, M.I. Transurethral resection of bladder tumour (TURBT). Transl. Androl. Urol. 2020,9, 3056–3072. [CrossRef]
9.
Song, P.H.; Cho, S.; Ko, Y.H. Decision based on narrow band imaging cystoscopy without a referential normal standard rather
increases unnecessary biopsy in detection of recurrent bladder urothelial carcinoma early after intravesical instillation. Cancer Res.
Treat. 2016,48, 273–280. [CrossRef]
10. Nagaya, T.; Nakamura, Y.A.; Choyke, P.L.; Kobayashi, H. Fluorescence-Guided Surgery. Front. Oncol. 2017,7, 314. [CrossRef]
11.
Cui, S.; Zhang, S.; Yue, S. Raman Spectroscopy and Imaging for Cancer Diagnosis. J. Healthc. Eng.
2018
,2018, 8619342. [CrossRef]
12.
Desroches, J.; Jermyn, M.; Mok, K.; Lemieux-Leduc, C.; Mercier, J.; St-Arnaud, K.; Urmey, K.; Guiot, M.C.; Marple, E.;
Petrecca, K.; et al.
Characterization of a Raman spectroscopy probe system for intraoperative brain tissue classification.
Biomed. Opt. Express 2015,6, 2380–2397. [CrossRef]
13.
Benard, A.; Desmedt, C.; Smolina, M.; Szternfeld, P.; Verdonck, M.; Rouas, G.; Kheddoumi, N.; Rothé, F.; Larsimont, D.;
Sotiriou, C.; et al.
Infrared imaging in breast cancer: Automated tissue component recognition and spectral characterization of
breast cancer cells as well as the tumor microenvironment. Analyst 2014,139, 1044–1056. [CrossRef] [PubMed]
14.
Old, O.J.; Fullwood, L.M.; Scott, R.; Lloyd, G.R.; Almond, L.M.; Shepherd, N.A.; Stone, N.; Barr, H.; Kendall, C. Vibrational
spectroscopy for cancer diagnostics. Anal. Methods 2014,6, 3901–3917. [CrossRef]
15.
Fernandez, D.C.; Bhargava, R.; Hewitt, S.M.; Levin, I.W. Infrared spectroscopic imaging for histopathologic recognition.
Nat. Biotechnol. 2005,23, 469–474. [CrossRef] [PubMed]
16.
Goormaghtigh, E. Infrared Imaging in Histopathology: Is a Unified Approach Possible? Biomed. Spectrosc. Imaging
2016
,5,
325–346. [CrossRef]
17.
Bhargava, R.; Madabhushi, A. Emerging Themes in Image Informatics and Molecular Analysis for Digital Pathology. Annu. Rev.
Biomed. Eng. 2016,18, 387–412. [CrossRef]
18.
Uckermann, O.; Juratli, T.A.; Galli, R.; Conde, M.; Wiedemuth, R.; Krex, D.; Geiger, K.; Temme, A.; Schackert, G.; Koch, E.; et al.
Optical Analysis of Glioma: Fourier-Transform Infrared Spectroscopy Reveals the IDH1 Mutation Status. Clin. Cancer Res.
2018
,
24, 2530–2538. [CrossRef]
Cancers 2023,15, 499 11 of 11
19.
Notarstefano, V.; Sabbatini, S.; Conti, C.; Pisani, M.; Astolfi, P.; Pro, C.; Rubini, C.; Vaccari, L.; Giorgini, E. Investigation of human
pancreatic cancer tissues by Fourier Transform Infrared Hyperspectral Imaging. J. Biophotonics 2020,13, e201960071. [CrossRef]
20.
Pezzei, C.; Brunner, A.; Bonn, G.K.; Huck, C.W. Fourier transform infrared imaging analysis in discrimination studies of bladder
cancer. Analyst 2013,138, 5719–5725. [CrossRef]
21.
Hughes, C.; Iqbal-Wahid, J.; Brown, M.; Shanks, J.H.; Eustace, A.; Denley, H.; Hoskin, P.J.; West, C.; Clarke, N.W.; Gardner, P.
FTIR microspectroscopy of selected rare diverse sub-variants of carcinoma of the urinary bladder. J. Biophotonics
2013
,6, 73–87.
[CrossRef]
22.
Kujdowicz, M.; Placha, W.; Mech, B.; Chrabaszcz, K.; Oko´n, K.; Malek, K. In Vitro Spectroscopy-Based Profiling of Urothelial
Carcinoma: A Fourier Transform Infrared and Raman Imaging Study. Cancers 2021,13, 123. [CrossRef]
23.
Sablinskas, V.; Bandzeviciute, R.; Velicka, M.; Ceponkus, J.; Urboniene, V.; Jankevicius, F.; Laurinaviˇcius, A.; Daseviˇcius, D.;
Steiner, G. Fiber attenuated total reflection infrared spectroscopy of kidney tissue during live surgery. J. Biophotonics
2020
,13,
e202000018. [CrossRef] [PubMed]
24.
Bandzeviciute, R.; Ceponkus, J.; Velicka, M.; Urboniene, V.; Jankevicius, F.; Zelvys, A.; Steiner, G.; Sablinskas, V. Fiber based
infrared spectroscopy of cancer tissues. J. Mol. Struct. 2020,1220, 128724. [CrossRef]
25.
Zacharovas, E.; Veliˇcka, M.; Platkeviˇcius, G.; ˇ
Cekauskas, A.; Želvys, A.; Niaura, G.; Šablinskas, V. Toward a SERS Diagnostic Tool
for Discrimination between Cancerous and Normal Bladder Tissues via Analysis of the Extracellular Fluid. ACS Omega
2022
,7,
10539–10549. [CrossRef]
26.
MedCalc Statistical Calculators. Available online: https://www.medcalc.org/calc/diagnostic_test.php?fbclid=IwAR0P2pB8
WaTor3OK_lYBg6jPzkMCz0X2obitS-sVYa3VwHsjXx7YedsiKMs (accessed on 22 December 2022).
27.
Talari, A.; Martinez, M.G.; Movasaghi, Z.; Rehman, S.; Rehman, I. Advances in Fourier transform infrared (FTIR) spectroscopy of
biological tissues. Appl. Spectrosc. Rev. 2017,52, 456–506. [CrossRef]
28.
Belbachir, K.; Noreen, R.; Gouspillou, G.; Petibois, C. Collagen types analysis and differentiation by FTIR spectroscopy.
Anal. Bioanal. Chem. 2009,395, 829–837. [CrossRef]
29.
Stevenson, K.K.; Kucich, U.; Whitbeck, C.; Levin, R.M.; Howard, P.S. Functional changes in bladder tissue from type III
collagen-deficient mice. Mol. Cell Biochem. 2006,283, 107–114. [CrossRef]
30.
Baba, A.; Câtoi, C. Chapter 3: Tumor Cell Morhology. In Comparative Oncology; The Publishing House of the Romanian Academy:
Bucharest, Romania, 2007. Available online: https://www.ncbi.nlm.nih.gov/books/NBK9557/ (accessed on 14 September 2022).
31.
Bannasch, P. Sequential cellular changes during chemical carcinogenesis. J. Cancer Res. Clin. Oncol.
1984
,108, 11–22. [CrossRef]
[PubMed]
32. Zois, C.E.; Favaro, E.; Harris, A.L. Glycogen metabolism in cancer. Biochem. Pharmacol. 2014,92, 3–11. [CrossRef]
33.
Mowatt, G.; N’Dow, J.; Vale, L.; Nabi, G.; Boachie, C.; Cook, J.A.; Fraser, C.; Griffiths, T.R.; Aberdeen Technology Assessment
Review (TAR) Group. Photodynamic diagnosis of bladder cancer compared with white light cystoscopy: Systematic review and
meta-analysis. Int. J. Technol. Assess. Health Care 2011,27, 3–10. [CrossRef] [PubMed]
34.
Babjuk, M.; Burger, M.; Compérat, E.M.; Gontero, P.; Mostafid, A.H.; Palou, J.; van Rhijn, B.W.G.; Rouprêt, M.; Shariat, S.F.;
Sylvester, R.; et al. European Association of Urology Guidelines on Non-muscle-invasive Bladder Cancer (TaT1 and Carcinoma In
Situ)—2019 Update. Eur. Urol. 2019,76, 639–657. [CrossRef] [PubMed]
Disclaimer/Publisher’s Note:
The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.
Available via license: CC BY 4.0
Content may be subject to copyright.
