High Sensitive Long Period Fiber Grating Biosensor for Cancer

Biomarker Detection

Giuseppe Quero

, Marco Consales

, Renato Severino

, Patrizio Vaiano

, Alessandra Boniello

Annamaria Sandomenico

, Menotti Ruvo

, Anna Borriello

, Laura Diodato

, Simona Zuppolini

Michele Giordano

, Immacolata Cristina Nettore

, Annamaria Colao

, Paolo Emidio Macchia

Flavio Santorelli

, Antonello Cutolo

and Andrea Cusano

Optoelectronics Group, Dept. of Engineering, University of Sannio, Benevento, Italy

Istituto di Biostrutture e Bioimmagini, Consiglio Nazionale delle Ricerche (IBB-CNR) and

Centro Interuniversitario di Ricerca sui Peptidi Bioattivi (CIRPeB), Napoli, Italy

Institute for Polymers, Composites and Biomaterials (IPCB) -CNR, Portici, Italy

Department of Clinical Medicine and Surgery, University of Napoli “Federico II”, Napoli, Italy

Hospital Consulting SpA, Bagno a Ripoli, Firenze, Italy

Keywords: Long Period Fiber Grating (

LPG), Optical Fiber Biosensor, Reflection Type LPG, Thyroglobulin.

Abstract: We report an innovative fiber optic biosensor based on Long Period Gratings (LPGs) working in reflection

configuration for real time monitoring of human Thyroglobulin (Tg), a protein marker of differentiated

thyroid cancer. A standard LPG is first transformed in a practical probe working in reflection mode, and

then it is coated with a single layer of atactic polystyrene (aPS) in order to increase its surrounding

refractive index sensitivity and to provide, at the same time, the desired interfacial properties for a stable

anti-Tg antibody. The functionalized reflection-type LPG biosensor clearly demonstrates the effectiveness

and sensitivity of the developed biosensing platform, allowing the real time and label-free detection of Tg in

the needle washouts of fine-needle aspiration biopsies, at concentrations useful for pre- and post-operative

assessment of the biomarker levels. Analyte recognition and capture were confirmed with a parallel on fiber

ELISA-like assay using, in pilot tests, the biotinylated protein and HRP-labeled streptavidin for its

detection. Dose-dependent experiments showed that the detection is linearly dependent on concentration

within the range between 0 and 4 ng/mL, while antibody saturation occurs for higher protein levels.

1 INTRODUCTION

The ever increasing incidence of cancer diseases is

imposing the development of highly sensitive and

effective tools for the real-time detection of

associated biomarkers for early diagnosis and

optimal treatment. This is particularly needed for the

diagnosis of papillary thyroid cancer, whose

incidence has dramatically increased over the past

few years in the United States and is predicted to

increase in the next years, recording a greater

frequency in the female population (Weir et al.,

2015). Papillary thyroid cancer is the most common

malignancy of the thyroid. Although it has favorable

long-term outcome, an early stage diagnosis is

fundamental to reduce morbidity and mortality. The

frequency of lymph nodes involvement is 27% to

46% at initial diagnosis and the recurrence rate is

3% to 30% during post-operative follow-up.

Distinguishing lymph nodes metastasis from benign

reactive lymphadenitis is therefore critical to rank

the malignancy risks in patients with papillary

thyroid cancer (Moon et al., 2013).

Thyroglobulin (Tg) is a thyroid specific 660 kDa

dimeric protein used by thyroid follicular cells as

precursor for biosynthesis of thyroid hormones.

Serum Tg levels are elevated in patients with goiter

and in several other clinical conditions, and to date,

the measurement of Tg is the mainstay in the post-

surgical follow-up of differentiated thyroid cancer

Quero, G., Consales, M., Severino, R., Vaiano, P., Boniello, A., Sandomenico, A., Ruvo, M., Borriello, A., Diodato, L., Zuppolini, S., Giordano, M., Nettore, I., Colao, A., Macchia, P., Santorelli,

F., Cutolo, A. and Cusano, A.

High Sensitive Long Period Fiber Grating Biosensor for Cancer Biomarker Detection.

DOI: 10.5220/0005846705610569

In Proceedings of the 9th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2016) - Volume 5: HEALTHINF, pages 561-569

ISBN: 978-989-758-170-0

561

(Pacini and Pinchera, 1999), As thyroid-specific

protein, its levels in lymph nodes are normally very

low and an increased Tg level in the needle washout

has been associated with metastasis of lymph nodes

in patients affected by differentiated thyroid

carcinoma (Giovanella et al., 2013). Its

determination is currently based on immunometric-

chemiluminescent or radioimmunometric assays

(Spencer and Lopresti, 2008).

In recent years, efforts to define and optimize

diagnostic and biosensing tools that incorporate such

features are significantly increased. Molecular

biosensors are preferred as clinical diagnostic tools

than other traditional methods because of real-time

measurement, rapid diagnosis, multi-target analyses,

automation, and reduced costs. A few works have so

far been proposed regarding Tg detection using a

biosensor platform. In 2008 Choi et al. detected Tg

in a cocktailed mixture of proteins by using the

competitive protein adsorption/exchange reactions,

namely Vroman effect. Implemented on a

microfluidic system, the target protein displaced a

pre-adsorbed weak-affinity protein on one surface of

the device, while another pre-adsorbed high-affinity

protein on an adjacent surface was not displaced.

Differential measurement using surface plasmon

resonance (SPR) phenomenon allowed Tg detection

(Choi and Chae, 2009). Recently Dantham et al.

(Dantham et al., 2013) reported the detection of

single human Tg protein molecule from the

resonance frequency shift of a whispering gallery

mode-nanoshell hybrid resonator upon adsorption on

the nanoshell. However, although the high

sensitivity of the proposed devices, the absence of a

bioreceptor featuring high specificity and affinity,

which can then discriminate between target and non-

target molecules, prevented the use of such systems

in clinical and diagnostic applications.

Moreover in the last years, the continuous

demand for lower limits of detection combined with

cost effectiveness and reliability features has been

the driving force for the successful demonstration of

optical label free biosensors with impressive figures

of merit (Fan et al., 2008; Hoa et al., 2007). Relative

principles of operation include SPR (Chung et al.,

2006; Teramura and Iwata, 2007), interferometry

(Weisser et al., 1999; Schneider et al., 1997), optical

waveguide-based biosensors (Website, http://

www.neosensors.com), optical ring resonators (Chao

et al., 2006; Ren et al., 2007; Hanumegowda et al.,

2005), fiber-based biosensors (Lee and Fauchet,

2007; Skivesen et al., 2007; Chryssis et al., 2005;

DeLisa et al., 2000; Zhang et al., 2005). Among the

others, fiber optic optrodes constitute a valuable

platform for label biosensing because of its intrinsic

biocompatibility, compact size, multiplexing

capability, remote operation and easy integration in

medical needles. In particular, in this work we

selected optical fiber LPGs as evanescent wave-

based biosensors for the measurements of local

refractive changes due to molecular binding

occurring at the sensor surface (Pilla et al., 2011;

Pilla et al., 2012, Del Villar et al., 2005; Cusano et

al., 2006, Pilla et al., 2009).

An LPG consists of a periodic modulation of the

refractive index (RI) at the core of an optical fiber

that results in the coupling of the light between core

and cladding modes (James and Tatam, 2003).

Thanks to the giant sensitivity to surrounding

refractive index (SRI) changes, LPGs represent a

very promising technological platform, which can be

employed in a wide number of chemical and

biological applications (Chiavaioli et al., 2014;

Eftimov, 2010; Baldini et al., 2012; Tripathi et al.,

2012; Smietana et al., 2015; Falciai et al., 2001;

Falate et al., 2005; Chen et al., 2007; Ramachandran

et al., 2002).

One of the peculiarities of LPGs platforms is

their operation in transmission mode which makes

the device sensitive to the bending, therefore

requiring the development of appropriate strain-free

packages to host the LPG device. In particular, the

bending applied on the LPG can introduce

unexpected variations in the spectrum of the

transmitted optical signal, thus complicating the

sensor signal interpretation. In addition the

capability to work in reflection mode addresses the

mentioned issues and allows the easy integration of

the reflective optrode in the vials containing the

biological solution and represents a more practical

and robust solution to be employed for concrete

biological applications (Huang et al., 2013; Quero et

al., 2015; Alwis et al., 2013; Cao et al., 2013).

In this work, we present the development of a

reflection-type LPG biosensor able to perform the

real time detection of thyroid cancer biomarkers in

the needle washouts of fine-needle aspiration

biopsies. After fabrication, the reflection-type LPG

is functionalized with a hydrophobic coating of a

specific bioreceptor, in our case an anti-Tg

monoclonal antibody and the protein is detected in

label-free experiments. Results clearly demonstrate

the effectiveness and sensibility of the biosensing

platform, allowing the in vitro detection of sub

ng/ml concentrations of human purified Tg. To

validate the potential translation of such LPG-based

biosensor into the clinical practice, detection

experiments on clinical samples have been carried out.

SmartMedDev 2016 - Special Session on Smart Medical Devices - From Lab to Clinical Practice

562

2 FABRICATION PROCESS

2.1 Reflection-type LPG Transducer

Fabrication

A customized LPG (Λ=370μm) UV-written in a

standard single-fiber was used for the biosensor

fabrication. To realize a more practical probe which

has to be immersed into laboratory vials containing

the biological samples under test, the crucial step is

the development of an LPG working in reflection

configuration. The first steps regarding the

fabrication of the reflection-type LPG (RT-LPG)

probe is the cut of the fiber inside which the LPG is

inscribed and the integration of a completely

reflecting layer (i.e. a mirror) on the fiber facet (see

Figure 1). A key aspect of this step is the

identification of the precise LPG position within the

fiber. This is of primary importance in order to cut

the fiber just after the grating, and to avoid the

formation of interference fringes within the

attenuation bands, typical of self-interfering LPGs

(Alwis et al., 2013).

Once identified the LPG position, a high

precision fiber cleaver (Fujikura CT-30) was used to

cut the optical fiber just after the grating, followed

by the integration of an Ag reflecting layer on the

facet of the cut fiber. To this aim, a silver mirror

reaction (Tollen’s test) was adopted (Yin et al.,

2002).

1620 1630 1640 1650 1660 1670

-35

-30

-25

-20

-15

-10

-5

Wavelength [nm]

Reflectance [dB]

Transmission spectrum

Reflection before mirror formation

Reflection after mirror formation

LPG

Cut position

Reflecting layer

(a)

Figure 1: LPG spectra acquired in air before the fiber cut

(blue curve), just after the fiber cut (red curve) and after

the mirror integration (green curve) and schematics of the

reflection-type LPG fabrication steps (in the insets).

Figure 1 shows the LPG spectra acquired just

before the optical fiber cut (i.e. in transmission

configuration), after the fiber cut (i.e. in reflectance

configuration before the mirror formation) and after

the mirror formation at the fiber end-face. It can be

seen that a significant baseline reduction occurs after

the cut (red curve), mainly due to the fact that light

passing through the LPG is mostly transmitted at the

fiber/air interface, and only a small portion (~3-4%)

of it is reflected back into the fiber. Nevertheless, as

soon as the Ag layer is formed at the fiber

termination, almost all the initial power is recovered

(green curve). We point out that the use of

reflection-type LPGs not only is of fundamental

importance to transform an LPG-based sensor in a

more practical probe for concrete biomedical

applications, but also improves the resonance

visibility (see Figure 1) due to the double passing of

light through the grating.

2.2 Overlay Deposition

The last fabrication step relied on the aPS overlay

deposition onto the RT-LPG surface using the dip-

coating (DC) technique. In particular, the aPS

overlay deposition was performed via the DC

process by means of an automated system (NIMA

Technology Micro-Processor Interface IU 4) at an

immersion/extraction speed of 100 mm/min (Pilla et

al., 2011; Pilla et al., 2009).

1580 1600 1620 1640 1660 1680 1700

-25

-20

-15

-10

-5

Wavelength [nm]

Reflectance [dB]

LPG bare

LPG aPS-coated

(a)

Figure 2: Spectra of the bare (blue line) and aPS-coated

(red line) RT-LPG.

Figure 2 reports the spectral position of the sixth

order cladding mode in the bare RT-LPG (blue

curve) and after the aPS overlay deposition (red

curve). All the spectra were recorded with the device

surrounded by air.

The aPS overlay thickness was chosen on the

basis of a preventive design process (carried out by

means of a virtual environment for the analysis and

simulation of nano-scale coated LPGs) in order to

optimize the device sensitivity in correspondence of

an SRI=1.340, which approaches the RI of the buffer

solution used in our binding experiments. The

optimized thickness overlay allows the modal

transition phenomena to take place and makes the

reflection-type LPG very high sensitive for a

SRI=1.340 is 310 nm.

High Sensitive Long Period Fiber Grating Biosensor for Cancer Biomarker Detection

563

The spectral characterization of the fabricated

LPG versus SRI has been carried out by submerging

the probe into aqueous glycerol solutions

characterized by different RI in the range 1.335-

1.460, in order to validate the fabrication process

success and the SRI sensitivity in correspondence of

an SRI=1.340. To this aim, an optoelectronic set-up

(see Figure 3) comprising a broadband light source

(with bandwidth 1200-1700 nm), a 2x1 directional

coupler and an optical spectrum analyzer (OSA,

ANDO AQ6317C, wavelength resolution 10 pm,

dynamic range 60 dB) was used for the acquisition

of the LPG reflection spectrum at the different

stages of the device fabrication and characterization.

The OSA is connected to a personal computer and

controlled by a LabView plug-in, enabling the

automatic acquisition of the RT-LPG spectra.

Acquired spectra are then automatically filtered and

elaborated by a MATLAB script that provides the

central resonance wavelengths (λc) of each

spectrum.

ANALIZZATORE DI

SPETTROOTTICO

(OSA)

SORGENTE ABANDA LARGA

DUT

FIBEROPTICCOUPLER

2X1

Figure 3: Schematic interrogation set-up.

During the biological experiments the RT-LPG

spectra is automatically acquired every 45 seconds,

thus providing a continuous and real time

monitoring of the interaction kinetics of the

biological molecules on the RT-LPG surface.

2.3 Surrounding Refractive Index

(SRI) Sensitivity Characterization

The spectral characterization of the fabricated LPG

versus SRI has been carried out by submerging the

probe into aqueous glycerol solutions characterized

by different RI in the range 1.335-1.460.

As reported in Figure 4b, the obtained SRI

sensitivity (|∂λc/∂SRI|) exhibits the typical

resonance-shaped behavior of transition mode LPG,

thus confirming the fabrication process success. At

the same time, it can be seen that the SRI sensitivity

in correspondence of an SRI=1.340 is equals to

~1700 nm/RIU.

1 1.1 1.2 1.3 1.4

500

1000

1500

2000

SRI

SRI Sensitivity [nm/RIU]

(

Opticalfiber

aPS layer

(c)

(a)

(b)

Figure 4: a) Schematic view of the final RT-LPG

transducer and b) Experimental SRI sensitivity

(|∂λc/∂SRI|) vs. SRI curve for the cladding mode LP

3 CHARACTERIZATION OF THE

TG/ANTI-TG MAB

INTERACTION

As a Tg bioreceptor we choose one of the several

commercially available anti-Tg monoclonal

antibodies because such reagents are generally

characterized by extremely high specificity and

affinity for the target molecules. The affinity of the

monoclonal antibody for Tg was however assessed

using both traditional immunoenzymatic and real-

time assays. An indirect ELISA was firstly

performed by adsorbing Tg on multiwell plates and

adding increasing amounts of the mouse anti-Tg

monoclonal antibody (Figure 5a). Following the

detection step with an anti-mouse antibody

conjugated with HRP, increasing chromogenic

signals (corresponding to increased bound analyte)

were recorded with increasing antibody

concentration. As shown in Figure 5a, a strong and

saturable binding signal was observed even at low

antibody concentrations. Saturation started at 100

pM mAb. Fitting of data points by a non-linear

algorithm, where we assumed a 1:1 binding

stoichiometry, provided a Dissociation constant

(KD) of 72 pM, a value indicative of the high

affinity of the antibody for the specific analyte. The

presence of high concentrations of BSA in the assay

(1% w/w, that is 10 mg/mL), also provided a very

strong indication of selectivity. Indeed blank signals

SmartMedDev 2016 - Special Session on Smart Medical Devices - From Lab to Clinical Practice

564

were negligible and after subtraction a 2.0 AU

residual signal was measured at saturation. To

further characterize the binding between the

antibody and Tg, we next performed a label free

binding assay using an SPR-based instrument

(BIAcore). The anti-Tg antibody and Tg were

thereby immobilized on the surface of the sensor

chip (1120 and 650 RU immobilization level,

respectively. Not shown). In one experiment with

the Tg-derivatized chip, mAb solutions at increasing

concentrations, between 0.25 and 1.5 nM, were

injected.

0 200 400 600 800 1000

0.0

0.5

1.0

1.5

2.0

anti-Tg mAb, pM

Abs

450nm

(a)

(b)

Figure 5: Biochemical system characterization. a) Dose-

response indirect ELISA assay performed to assess

antigen/antibody affinity. b) SPR assay using Biacore of

the binding kinetic of anti-Tg monoclonal antibody to

immobilized Tg, at different antibody concentrations.

As shown in Figure 5b, dose-response association

and dissociation curves were obtained witnessing the

high affinity and specificity of the interaction. Using

the kinetic parameters a KD of 70 pM was

estimated. This value was in full agreement with that

extrapolated by ELISA, confirming the strength of

the interaction and the high specificity. When we

probed the immobilized antibody with soluble Tg a

significantly higher KD (about 1.3 nM) was

determined (data not shown) than that determined by

ELISA. This was likely due to inappropriate

antibody immobilization.

3.1 Assessment of Tg Capture on LPG

Biosensor Surface by on-Fiber

ELISA-like Assay

Figure 6a shows a typical sensorgram, reporting the

measured resonance wavelenghts as function of

time, observed for a biotinylated human

Thyroglobulin detection assay and in the inset the

magnification of the experiment final step regarding

the Tg detection.

(a)

(b)

(c)

Figure 6: a) Sensorgram obtained for the detection of

40µg/ml biotinylated human Thyroglobulin using the

reflection-type Long Period Grating biosensing platform;

b) schematic view of the ELISA-like assay performed on

the functionalized LPG; c) binding signals obtained by the

ELISA-like assay.

After a prolonged and satisfactory adaptation of

the functionalized biosensor in the running buffer,

an anti-Thyroglobulin monoclonal antibody,

previously selected for its affinity and specificity for

the identified marker and deeply characterized by

label-free and immunoenzymatic assays, was

immobilized on the probe active surface via

hydrophobic coating, recording a 1.9 nm wavelength

shift of the resonant peak.

After analyte capture to the specific receptor (see

magnification in Figure 6a, a 2.0 nm wavelength

shift of the resonant peak was observed. To avoid

the influence of the different solutions refractive

index change, peak shifts have been estimated as a

difference between two successive immersions in

the running buffer, specifically before and after

dipping the sensor in the test solution. In addition,

by exploiting the binding of HRP-labeled

streptavidin to the biotinylated Thyroglobulin,

analyte capture was confirmed in a parallel on fiber

ELISA-like assay (Figure 6b-c).

High Sensitive Long Period Fiber Grating Biosensor for Cancer Biomarker Detection

565

4 DETECTION OF TG WITH LPG

BIOSENSOR

4.1 Calibration Curve for the Detection

of Human Tg with LPG Biosensor

To use the platform for the detection of Tg in

biological fluids, we first evaluated the system dose-

response features and the associated sensitivity. The

system was therefore tested on a set of several

different Tg solutions at different concentrations

ranging between 0.08 ng/mL (0.13 pM) to 88 ng/mL

(146 pM). Since no regeneration steps were carried

out, different LPG biosensors were used for this test

and each transducer was used for the detection of

two cumulative and consecutive analyte

concentrations.

At the different concentrations (0.08 ng/mL, 0.88

ng/mL, 4.0 ng/mL, 8.0 ng/mL, 44 ng/mL and 88

ng/mL), we observed the following average values

of Δλ

Tg-binding

: 0.26 nm, 1.03 nm, 1.65 nm, 2.45 nm,

3.35 nm and 3.58 nm, respectively. A blank average

value of 0.10 nm, accounting for the non-specific

binding, was instead measured using the transducer

without antibody coating. The corresponding dose-

dependent curve is reported in Figure 7, where we

can observe the dose-dependency and the saturation

reached yet between about 20 and 30 ng/mL of Tg

(33 pM and 50 pM, respectively). Also, the analyte

could be still detected with a remarkable difference

over the non-specific recognition at the lowest

concentration of 0.08 ng/mL.

Although the fabrication process was robust and

repeatable, a normalization procedure was needed to

account for the even tiny differences of device

performances. Such differences may occur due to

the LPG fabrication tolerances, or to slightly

different antibody coatings obtained on the distinct

devices. To normalize the signals we used the

wavelength shift occurred upon mAb coating, as an

indirect measure of surface sensitivity of each LPG

probe. We thus calculated the final observable (O) as

reported in the following equation 1:

Observable (O) = Δλ

indin

+ Δλ

mAb-coatin

(1)

Where, Δλ

Tg-binding

and Δλ

mAb-coating

are the

differences between the stabilized central

wavelength of the attenuation band in buffer

solutions before and after contact with the

biomolecule solution and denote Tg binding and

mAb coating onto the LPG surface, respectively.

The KD was determined by reporting the observed

resonance peak shift at every concentration versus

analyte concentration. Some antibody-coated optical

fibers were immersed in buffer alone and used as

blank.

Hydrophobic adsorption of the antibody can be

influenced by several factors, such as temperature,

exposure time, different surface properties and

ligand orientation, thus influencing transducer

sensitivity. Such variability could result in different

effective detection capability and sensitivity between

different functionalized biosensors. Moreover, any

sensitivity variation of the biosensor associated to

different polymer thickness, that could cause

significant underestimation of Tg concentration, is in

this way corrected. It is also important to underscore

that the dip-coating technique used to deposit the

polymer on the optical fibers, does not guarantee a

tight control over polymer thickness at the

nanometer scale. The normalization that we have

introduced attenuates the impact of such factors and

we indeed observed an optimal correlation between

data obtained with different optical fibers.

0 50 100

0.2

0.4

0.6

0.8

Tg Concentration [ng/mL]

Observable O [nm/nm]

h-Tg

fitting data

Figure 7: Calibration curve for the semi-quantitative

detection of human Thyroglobulin using RT-LPG

biosensor. Blue dots refer to the dose-dependent assay

performed to obtain the calibration curve.

Data clearly show that human Tg bound with

high affinity and in a dose dependent manner to the

monoclonal antibody immobilized onto the solid

phase. Also, it is worth noting that the level of

nonspecific adsorption of analyte to the surface is

particularly low, suggesting that functionalization

and blocking were particularly effective and that

selectivity was also particularly high, given the very

poor recognition of albumin present at high

concentration.

From the plot of resonance peak shifts against

analyte concentration reported in Figure 7, we also

estimated a KD of about 6 pM, which is in the same

low pM range of that determined by ELISA and by

Biacore (about 70 pM). Such KD value also reflects

the high sensitivity of the detection system, which is

able to detect as low as 0.08 ng/mL Tg under these

SmartMedDev 2016 - Special Session on Smart Medical Devices - From Lab to Clinical Practice

566

conditions. A linear dose-response is grossly

obtained within the 0 – 4 ng/mL range (0 – 6.7 pM).

4.2 Detection of Tg in Clinical Samples

RT-LPG-based biosensors were used for the ex-vivo

detection of human Tg from needle washouts of

fine-needle aspiration biopsies of thyroid nodules

from several different patients.

(a)(a)

0 50 100 150 200 250 300 350 400

1530

1540

1550

1560

1570

1580

1590

1600

1610

Time [min]



[nm]

III

Running

Buffer

BSA

Running

Buffer

VII

Running

Buffer

Running

Buffer

Mono Ab

260 280 300 320 340 360 380

1545

1550

1555

1560

1565

1570

1575

Time [min]



[nm]

Running

Buffer

VII

Running

Buffer

Tg from fine-needle

aspiration biopsy

Figure 8: RT-LPG sensorgram reporting the wavelength

shift of 6th order cladding mode attenuation band during

human Thyrogobulin detection from the needle washout of

fine-needle aspiration biopsy. In magnification the

wavelength shift during Thyroglobulin binding event is

reported.

Figure 8 shows typical sensorgrams obtained for

such the real-time detection of human Tg. The

transducer was functionalized as described above by

immersing the probe in a solution containing the

anti-Tg mAb. After the blocking step necessary to

reduce non-specific adsorption of the analyte, the

probe was immersed in solutions obtained by

diluting the needle washout of fine-needle aspiration

biopsies (steps V-VII in the magnification). Such

solutions contained Tg at different concentrations,

previously quantified using a standard

immunoenzymatic assay. Different independent

assays were performed with solution samples at

increasing Tg concentrations, observing a

remarkable correlation with the calibration curve

previously obtained (Figure 9, red dots) and with the

concentrations obtained by canonical Tg

quantification. In particular, a first set of human

samples was diluted at 1 ng/mL Tg (within the

linearly responsive range) and detection was

performed with different RT-LPG biosensors. As

shown in Figure 9, we obtained an average value for

the observable O of 0.225 corresponding to 1.45

ng/mL Tg, a value in very good agreement with the

expected concentration. An independent detection

assays were performed on human samples prepared

at 5 ng/mL Tg, obtaining average O values of 0.430,

corresponding to 4.5 ng/mL. Also in this case values

were in an overall satisfactory agreement with those

expected, confirming the linear dose-response within

the expected range and, most importantly, the

substantial lack of strong non-specific interactions

with other high concentration plasma proteins that

contaminate the washout samples. Finally, the

detection assays were performed on undiluted

samples, coming from different patient specimens,

having Tg concentrations higher than 3000 ng/mL.

The average O values obtained after Tg capture were

0.777, corresponding, as expected, to complete

biosensor saturation. We reported the values on the

plot of Figure 9 as red dots after the dashed line to

underline the substantial accordance between Tg

detection with LPG biosensors and canonical

immunoenzymatic techniques.

Such reference values clearly indicate that the

biosensor operates at best in a range of

concentrations matching that having a high clinical

relevance.

0 50 100 140

0.2

0.4

0.6

0.8

Tg Concentration [ng/mL]

Observable O [nm/nm]

h-Tg

fitting data

human sample

Figure 9: Calibration curve for the semi-quantitative

detection of human Thyroglobulin using RT-LPG

biosensor and ex-vivo detection of human Thyroglobulin

from needle washout of fine-needle aspiration biopsy.

Blue dots refer to the dose-dependent assay performed to

obtain the calibration curve, while red dots refer to the ex-

vivo assay performed on human samples. The response is

roughly linear within the range 0 – 4 ng/mL.

In order to improve prognosis of differentiated

thyroid carcinomas, it is widely accepted that

confirmed or suspected cervical Lymph Nodes (LN)

metastases should be removed for local control.

However either preoperatively with ultrasonography

(US) or during the operation, it is difficult to

recognize small and occult LN metastases in the

central compartment (Chéreau et al., 2015).

Although it is widely accepted that differentiated

thyroid carcinoma recurrences after lymph node

dissection is unrelated to the number of LNs

removed (Albuja-Cruz et al., 2012), it is also well

established that removing occult LN metastasis

decreases the rate of recurrence in the neck (Roh et

High Sensitive Long Period Fiber Grating Biosensor for Cancer Biomarker Detection

567

al., 2011). The clinical application of real time Tg

detection will probably in a near future provide to

surgeons a powerful and reliable tool to precisely

identify metastatic lymph nodes.

5 CONCLUSIONS

We report an innovative fiber optic nano-optrode

based on LPGs working in reflection mode for real

time monitoring of human Thyroglobulin, a protein

marker of differentiated thyroid cancer.

The reflection-type LPG biosensor, coated with a

single layer of atactic polystyrene onto which a

specific, high affinity anti-Tg antibody was

adsorbed, allowed the real time and label-free

detection of Tg in the needle washouts of fine-

needle aspiration biopsies, at concentrations useful

for pre- and post-operative assessment of the

biomarker levels.

Analyte recognition and capture were confirmed

with a parallel on fiber ELISA-like assay using, in

pilot tests, the biotinylated protein and HRP-labeled

streptavidin for its detection. Dose-dependent

experiments showed that the detection is linearly

dependent on concentration within the range

between 0 and 4 ng/mL, while antibody saturation

occurs for higher protein levels. The system is

characterized by a very high sensitivity and

specificity, which reflects the specificity and affinity

of the antibody chosen as capturing bioreceptor.

Indeed, the biosensor allowed the ex-vivo detection

of sub ng/ml concentrations of human Tg from

needle washouts of fine-needle aspiration biopsies of

thyroid nodule from different patients.

Data here presented underline the high potential

of the proposed biosensing platform and the several

advantages of this kind of transducer: the absence of

labeling requirements represents in fact an attractive

alternative to traditional label-based techniques such

as fluorescence, colorimetry or radioactivity-based

approaches, without affecting the intrinsic affinity.

Furthermore, it appears particularly useful for

monitoring post-operative Tg levels, as the Tg

warning concentration (about 1 ng/mL) is well

within our linear response range (about 0 – 4

ng/mL). We foresee that, following a further

optimization and standardization of the detection

protocol, such an application is close at hand under

the current experimental settings. We also believe

that a further engineering of the detection platform

could allow the detection of Tg during biopsy

collection. However this would require suitable

needles and proper microfluidic and washing

devices for the removal of tissue debris and for

analyte dilution, a step strongly needed in light of

the very high sensitivity of the detection system.

Although the detection times are still in the hour

range, we do believe that the integration with a

microfluidic system will significantly reduce the

response time, making this method time saving, an

essential parameter for its application in the clinical

field. In addition a multiplexed configuration, with

the parallel detection of several biomarkers of

clinical interest, appears as a possible future target.

ACKNOWLEDGEMENTS

The authors gratefully acknowledge the financial

support from the national project “Smart Health

2.0”, funded by the Italian Ministry of Education,

University and Research (MIUR) under the PON

framework.

REFERENCES

Albuja-Cruz, M. B., Thorson, C. M., Allan. B. J., Lew. J.

I., Rodgers. 2012, Surgery, 152, 1177–1183.

Alwis, L., Sun, T., Grattan, K. T. V., 2013. Sens.

Actuators, B, 178, 694–699.

Baldini, F., Brenci, M., Chiavaioli, F., Giannetti,

A.,Trono, C., 2012. Anal. Bioanal. Chem. 402, 109–

116.

Cao, J., Tu, M. H., Sun, T., Grattan, K. T. V., 2013. Sens.

Actuators, B, 181, 611–619.

Chao, C. Y., Fung, W., Guo, L. J., 2006. IEEE J. Sel. Top.

Quantum Electron. 12, 134-142.

Chen, X., Zhou, K., Zhang, L., Bennion, I., 2007. Appl.

Opt. 46, 451–455.

Chéreau, N., Buffet, C., Trésallet, C., Tissier, F.,

Leenhardt, L., Menegaux, F. 2015. Surgery. 15, 704-

707.

Chiavaioli, F., Biswas, P., Trono, C., Bandyopadhyay, S.,

Giannetti, A., Tombelli, S., Basumallick, N.,

Dasgupta, K., Baldini, F., 2014. Biosens. Bioelectron.

60, 305–310.

Choi, S., Chae, J., 2009. Biosens. Bioelectron. 25, 118–

123.

Chryssis, A. N., Saini, S.S., Lee, S. M., H. Yi, Bentley,

W.E., Dagenais, M., 2005. IEEE J. Sel. Top. Quantum

Electron. 11, 864-872.

Chung, J. W., Bernhardt, R., Pyun, J. C., 2006. Sens.

Actuators B Chem. 118, 28-32.

Cusano A., Iadicicco, A., Pilla, P., Contessa, L.,

Campopiano, S., Cutolo, A., Giordano, M., 2006a.

Opt. Express, 14, 19–34.

Dantham, V. R., Holler, S., Barbre, C., Keng, D.,

Kolchenko, V., Arnold, S., 2013. Nano Lett. 13, 3347-

3351.

SmartMedDev 2016 - Special Session on Smart Medical Devices - From Lab to Clinical Practice

568

DeLisa, M. P., Zhang, Z., Shiloach, M., Pilevar, S., Davis,

C.C., Sirkis, J.S., Bentley, W.E., 2000. Anal. Chem.

72, 2895–900.

Del Villar, I., Matías, I. R., Arregui, F. J., Lalanne, P.,

2005. Opt. Express, 13, 56–69.

Eftimov, T., 2010. Applications of Fiber Gratings in

Chemical and Biochemical Sensing, in: Zourob, M.,

Lakhtakia, A. (Eds.), Optical Guided-wave Chemical

and Biosensors II, Springer Series on Chemical

Sensors and Biosensors 8, Springer Berlin,

Heidelberg, pp. 151–176.

Falate, R., Kamikawachi, R. C., Müller, M., Kalinowski,

H. J., Fabris, J. L., 2005. Sens. Actuators B Chem. 105,

430–436.

Falciai, R., Mignani, A.G., Vannini, A., 2001. Sens.

Actuators B Chem. 74, 74–77.

Fan, X., White, I. M., Shopova, S. I., Zhu, H., Suter, J. D.,

Sun, Y., 2008. Anal. Chim. Acta. 620, 8-26.

Garg, R., Tripathi, S. M., Thyagarajan, K., Bock, W. J.,

2013. Sens. Actuators, B, 176, 1121–1127.

Giovanella, L., Bongiovanni, M., Trimboli, P., 2013. Curr

Opin Oncol., 25: 6-13.

Hanumegowda, N.M., White, I.M., Oveys, H., Fan, X.,

2005. Sens. Lett. 3, 315-319.

Hoa, X. D., Kirk, A.G., Tabrizian, M., 2007. Biosens.

Bioelectron. 23, 15–160.

Huang, J., Lan, X., Kaur, A., Wang, H., Yuan, L., Xiao,

H., 2013. Optical Engeneering, 52, 014404.

James, S. W., Tatam, R. P., 2003. Meas. Sci. Technol. 14,

R49–R61.

Lee, M. R., Fauchet, P. M., 2007. Opt. Express 15, 4530-

4535.

Moon, J. H., Yong I. K., Lim, J. A., Choi, H. S., Cho, S.

W., Kim, K. W., Park, H. J., Paeng, J. C., Park, Y. J.,

Yi, K. H., Park, D. J., Kim, S. E.and .Chung, J. K.,

2013. J Clin Endocrinol Metab, 98, 1061-1068.

Pacini, F., Pinchera, A., 1999. Biochimie., 81:463-467.

Patrick, H. J., Kersey, A. D., Bucholtz, F., 1998. J.

Lightwave Technol. 16, 1606–1612.

Pilla, P., Manzillo, P., Malachovská, V., Buosciolo, A.,

Campopiano, S., Cutolo, A., Ambrosio, L., Giordano,

M., Cusano, A., 2009. Opt. Express, 17, 20039–20050.

Pilla, P., Malachovská, V., Borriello, A., Buosciolo, A.,

Giordano, M., Ambrosio, L., Cutolo, A., Cusano, A.,

2011. Opt. Express, 19, 512–526.

Pilla, P., Sandomenico, A., Malachovská, V., Borriello,

A., Giordano, M., Cutolo, A., Ruvo, M., Cusano, A.,

2012. Biosens. Bioelectron. 31, 486–491.

Quero, G., Consales, M., Vaiano, P., Cusano, A.,

Zuppolini, S., Diodato, L., Borriello, A., Giordano,

M., Venturelli, A., Costi, M. P., 2015. XVIII AISEM

Annual Conference.

Ramachandran, S., Wang, Z., Yan, M., 2002.

Opt. Lett.

27, 698–700.

Ren, H. C., Vollmer, F., Arnold, S., Libchaber, A., 2007.

Opt. Express 15, 17410-17423.

Roh, J. L., Kim, J. M., Park, C. I. 2011. Ann Surg Oncol,

18, 1312–1318.

Schneider, B. H., Edwards, J. G., Hartman, N. F., 1997.

Clin. Chem. 43, 1757-1563.

Skivesen, N., Tetu, A., Kristensen, M., Kjems J., Frandsen

L. H., Borel P. I, 2007. Opt. Express 15, 3169-3176.

Smietana, M., Koba, M., Brzozowska, E., Krogulski, K.,

Nakonieczny, J., Wachnicki, L., Mikulic, P.,

Godlewski, M., Bock, W. J., 2015. Opt. Express. 23,

8441–8453.

Spencer, C. A., Lopresti, J. S., 2008. Nature Clinical

Practice 4, 223–233.

Teramura, Y., Iwata, H., 2007. Anal. Biochem. 365, 201-

207.

Tripathi, S. M., Bock, W. J., Mikulic, P., Chinnappan, R.,

Ng, A., Tolba, M., Zourob, M., 2012. Biosens.

Bioelectron. 35, 308–312.

Website, http://www.neosensors.com.

Weir, H. K., Thompson, T. D., Soman, A., Moller, B.,

Leadbetter, S., 2015. Cancer, 121, 1827-1837.

Weisser, M., Tovar, G., Mittler-Neher, Knoll, S. W.,

Brosinger, F., Freimuth, H., Lacher, M., Ehrfeld, W.,

1999. Biosens. Bioelectron. 14, 405-411.

Yin, Y., Li, Z. Y., Zhong Z., Gates B., Venkateswaran S.,

2002. S. J. Mater. Chem. 12, 522−527.

Zhang, Y., Shibru, H., Cooper, K.L., Wang, A., 2005. Opt.

Lett. 30, 1021–1023.

High Sensitive Long Period Fiber Grating Biosensor for Cancer Biomarker Detection

569