2.1.

Imaging Agents

(ICG)-NAVPNLRGDLQVLAQKVART was purchased from China Peptides Co. Ltd. The amino acid peptide chain reaction was performed based on 2-chlorotrityl chloride resin with reference to the classical solid-phase peptide synthesis method. Kaiser reagent was used for monitoring during the reaction to ensure the linkage efficiency of amino acids and the efficiency of protecting group removal. All the reactions were performed at room temperature. ICG-COOH modification was performed using a condensation reagent for NHS activation, followed by the addition of a catalyst for linking to the N-terminus of the peptide. The resin and excess protecting groups were removed using trifluoroacetic acid, residual impurities were removed by reverse high-performance liquid chromatography, and the target material was collected and lyophilized to obtain the final product. The probe was prepared at $1\mathrm{mg}/\mathrm{mL}$ concentration in 50% acetonitrile/50% ${\mathrm{H}}_2\mathrm{O}$.

2.2.

Clinical Tissue and Hematoxylin-and-Eosin Staining

Formalin-fixed, paraffin-embedded biospecimens were collected from the Oral Surgery Department of the Ninth People’s Hospital. All diagnoses were made by pathologists through hematoxylin-and-eosin (HE)-stained sections. We analyzed specimens from four groups (normal [paraneoplastic tissue], $n=6$; benign [papilloma], $n=6$; precancerous [white spots], $n=6$; and malignant oral cancer tissue, $n=6$). All patients signed an informed consent form prior to surgery. The protocol number of the clinical research project of the Medical Ethics Committee was N0JBWKQA002.

2.3.

Histology and Immunohistochemistry

Tissue sections were cut and immunohistochemically (IHC) stained for diagnosis. The tissues were divided into four groups according to the malignancy grade. Sections were dewaxed with xylene and alcohol, repaired with an antigen repair solution, and washed with phosphate-buffered saline (PBS). The samples were blocked after drawing circles around them; subsequently, primary antibodies for $\alpha \mathrm{v}\beta 6$ and EGFR were added dropwise at a temperature of 4°C overnight. The primary antibody was recovered, washed, and placed in the PBS solution. Then, color development was performed with diaminobenzidine solution; this was followed by re-staining, alcohol separation, and blocking. Microscopic photography was performed, and the sections were observed using an Olympus CX43 Biological Microscope (Olympus, Japan).

2.4.

Quantification of $\alpha \mathrm{v}\beta 6$ and EGFR

For $\alpha \mathrm{v}\beta 6$ and EGFR protein quantification, IHC slides were digitalized using a Pannoramic DESK, P-MIDI, P250, and P1000 (3DHISTECH, Budapest, Hungary). In the surgical specimens, we separately analyzed tumor and normal tissues. Thereafter, $\alpha \mathrm{v}\beta 6$ and EGFR were quantified using Image J/FIJI software for at least three fields of view per area of interest (tumor and normal tissue) ($20\times$ magnification).

Diaminobenzidine and HE staining were separated using the color deconvolution algorithm, and appropriate threshold levels were set to measure the area of specific diaminobenzidine staining ($\alpha \mathrm{v}\beta 6$ area) and HE staining (tissue area) to calculate the relative $\alpha \mathrm{v}\beta 6$ positive area on each image. The thresholds were kept constant within each dataset. We recorded the mean and standard deviation for each sample. For grouped analysis, we used a paired $t$-test.

2.5.

Cell Lines and Reagents

Four cell lines were used in this study: CAL27, SCC9, FaDu, and oral normal germ cells (HIOECs). FaDu and CAL27 were procured from Procell Life Science & Technology Co., Ltd. SCC9 cells were purchased from Cobioer Biosciences Co., Ltd. Dulbecco’s modified eagle medium (DMEM) was obtained from Shanghai Zhi Bei Biotechnology Co., Ltd. FaDu cells were cultured in minimum essential medium. CAL27, SCC9, and HIOEC were cultured in DMEM. Cells were cultured in DMEM (Gibco, Invitrogen, Carlsbad, California) supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin at 37°C and 5% ${\mathrm{CO}}_2$. To ensure exponential growth, the cells were reseeded weekly.

2.6.

Mouse Xenograft and Inflammation Models

The School of Pharmacy, Shanghai Ninth People’s Hospital, Shanghai Jiao Tong University approved all animal studies (SH9H-2020-A620-1) and ensured that all procedures were performed in accordance with the National Institutes of Health standards for the care and use of laboratory animals. The mice were maintained on a 12-h light cycle and regularly fed and were divided into three groups (according to the mouse model and the imaging probe), as shown below:

Tongue cancer model: $n=6$ injections of NHS-ICG, $n=6$ injections of $\alpha \mathrm{v}\beta 6$-ICG.

Inflammation model: $n=6$ NHS-ICG injections, $n=6$ $\alpha \mathrm{v}\beta 6$-ICG injections.

Normal mice model: $n=6$ injected with NHS-ICG, $n=6$ injected with $\alpha \mathrm{v}\beta 6$-ICG.

2.7.

Tumor Xenograft Model

This model included NSG mice (half male and half female; 4 to 6 weeks old). Before xenografting, the cells were washed with PBS and re-suspended in a $1:1$ mixture of PBS and Matrigel matrix (Corning). For in vivo and ex vivo experiments, we grew in situ animal models using 6-week-old nude mice and human tongue squamous cell carcinoma cells (CAL27) by injecting $2\times {10}^7\mathrm{cells}$ in a mixture of culture medium ($200\mu \mathrm{L}$). The tumor injectate was infiltrated into the tongues of the mice. When the tumors grew to $50{\mathrm{mm}}^3$, we performed the experiments after xenografting. The duration of tumor growth was 1 week.

2.8.

Inflammation Model

This model included BALB/c mice (half male and half female; 4 to 6 weeks old). All BALB/c mice were acclimatized in the laboratory for 4 to 6 weeks, and a 5-mm long, 1-mm deep wound was cut vertically in the midline of the mouse tongue using a scalpel. Live imaging was performed $\sim 5$ to 7 days later.

2.9.

Reverse Transcription-Polymerase Chain Reaction

Total RNA was extracted from tissues or cells using TRIzol® reagent (Invitrogen; Thermo Fisher Scientific, Inc., Carlsbad, California) according to the manufacturer’s protocols. The extracted RNA was reverse-transcribed using the PrimeScript RT kit with gDNA Eraser (Takara Biomedical Technology Co., Ltd., Beijing, China) according to the manufacturer’s protocols. Quantitative polymerase chain reaction (PCR) was performed using the SYBR Green PCR Master Mix (Takara Biomedical Technology Co., Ltd.) according to the manufacturer’s protocol. Relative expression was normalized to endogenous glyceraldehyde 3-phosphate dehydrogenase (GAPDH). The forward and reverse primer sequences were as follows: AQP9 forward, 5′-AAATAAACCTCCTTGG CCTGA-3′ and reverse, 5′-GCAACAAACATCACCA CACC-3′; E-Cad forward, 5′-CAATGGTGTCCATG GAACA-3′ and reverse, 5′-CCTCCTACCCTCCTGTTCG-3′; α-SMA forward, 5′-TCCCTTGAGAAGAGTTAC GAGTTG-3′ and reverse, 5′-ATGATGCTGTTGTAGGT GGTTTC-3′; N-Cad forward, 5′-CAGTATCCGGTCC GATCTGC-3′ and reverse, 5′-GTCCTGCTCACCACCACTAC-3′; DVL2 forward, 5′-AGTCAGCTCT CATGTTGAGGGT-3′ and reverse, 5′-TACCCAGCCCACACCT TCTT-3′; GSK-3β forward, 5′-CCGACTAACACCA CTGGAAGCT-3′ and reverse, 5′-AGGATGGTAGCCA GAGGTGGAT-3′; CyclinD1 forward, 5′-GAGACCATC CCCCTGACGGC-3′ and reverse, 5′-TCTTCCTCCTCCTCGGCGGC-3′; β-catenin forward, 5′-TGCAGTTCGCCTTCACTATG-3′ and reverse, 5′-ACTAGTCGTGGAATGGCACC-3′; and GAPDH forward, 5′-GCAAGAGCACAAGAGGAAGA-3′ and reverse, 5′-ACTGTGAGGAGGGGAGATTC-3′. Relative expression levels were calculated using the $2^{-\mathrm{\Delta }\mathrm{\Delta }Cq}$ method.

2.10.

Western Blot

EGFR and $\alpha \mathrm{v}\beta 6$ protein expression levels were measured in CAL27, SCC9, FaDu, and HIOEC cell lysates using Western blot (WB) analysis. Proteins were isolated from cells, separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and transferred onto a nitrocellulose membrane. Proteins were detected using antibodies specific for $\alpha \mathrm{v}\beta 6$ (Abcam ab187155; 1:1000), EGFR (Abcam ab52894; 1:1000), and b-actin (LF202; 1:5000) with a corresponding horseradish peroxidase-conjugated secondary antibody (Abcam ab205718; 1:50,000). An enhanced chemiluminescent substrate (ECL Coloring Solution, Tianneng Company) was used for detection. The $\alpha \mathrm{v}\beta 6$ and EGFR protein expression levels were measured in clinical tissues using WB analysis as described above.

2.11.

Histology and Immunofluorescence

The tongues of the mice were harvested and cut into pieces after 1 week, and the tissues were classified according to the type of lesion. One part was fixed in 10% neutral-buffered formalin and embedded in paraffin for IHC analysis. The other part was frozen in an optimal cutting temperature embedding compound, and $30\text{-}\mu \mathrm{m}$ sections were cut for fluorescence imaging using Pannoramic DESK, P-MIDI, P250, and P1000 (3D HISTECH) at the 800-nm channel with $5\text{-}\mu \mathrm{m}$ resolution. For HE staining, serial $5\text{-}\mu \mathrm{m}$ sections were cut.

2.12.

Confocal Laser Scanning Microscopy

To select a certain type of cell and a suitable probe concentration for the best fluorescence imaging, we performed two experiments. The first experiment was performed using a concentration of $1\mu \mathrm{M}$. We performed the experiment in different cell lines (HIOEC as a normal contrast; the other types were CAL27, FaDu, and SCC9). We pre-inoculated six-well plates with each cell line, treated them with a concentration ($1\mu \mathrm{M}$) of fluorescent probes and trypsin-digested cells, and centrifuged them at $300g$ for 5 min. We collected the cells and re-suspended ${10}^5$ cells in $200\mu \mathrm{L}$ of binding buffer. We added $4\mu \mathrm{L}$ of $0.5\mathrm{mg}/\mathrm{mL}$ propidium iodide (PI) and $2\mu \mathrm{L}$ of Annexin V-FITC solution. The cells were then incubated in the dark for 15 min at room temperature, away from light. Fluorescence was detected using flow cytometry. The maximum excitation wavelength of Annexin V-FITC was 488 nm, and the emission wavelength was 520 nm. The maximum excitation wavelength of PI was 488 nm, and the emission wavelength was 561nm.

The second experiment was performed using CAL27. We used different concentrations of the fluorescent probes (such as 30, 100, 300, 1000, and 30,000 nM). These steps are similar to those described above.

2.13.

Cell Viability (CCK8)

The cells were divided into two groups: tumor cell lines (CAL27, FaDu, and SCC9) and a normal cell line (HIOEC). After each type of cell pre-inoculated in 96-well plates was treated with different concentrations of fluorescent probes, $10\mu \mathrm{L}$ CCK8 solution ($100\mu \mathrm{L}$ culture system) was added to each well. Wells with CCK8 and normal cells were used as controls. Cells were incubated in a cell incubator (37°C, 5% ${\mathrm{CO}}_2$) for 1 h. The absorbance was measured at 450 nm. The relative viability of the cells in each well was calculated based on the absorbance values.

2.14.

Data Acquisition and Analysis

Fluorescence was measured in radiant efficiency using Living Image, v4.0 (PerkinElmer, Inc., Waltham, Massachusetts) for Xenogeny-IVIS images and fluorescence counts using Image Studio, v3.0 (LI-COR, Inc., Lincoln, Nebraska) for Odyssey images. Statistical analyses were performed using GraphPad software.

Paired $t$-tests were used to quantitatively compare fluorescence, except where otherwise noted. Analysis of variance was used to compare fluorescence across several categorical groups (i.e., cell lines). Pearson’s correlation coefficient was used to analyze the relationship between ICG and $\alpha \mathrm{v}\beta 6$-ICG fluorescence. Statistical significance was defined as $p<0.05$.

3.1.

$\alpha \mathrm{v}\beta 6$ Is Overexpressed in the Developing Stages of HNSCC

To explore the clinical value of an $\alpha \mathrm{v}\beta 6$-targeted imaging approach in HNSCC, we quantified $\alpha \mathrm{v}\beta 6$ expression in human surgical biospecimens ($n=6$). In all specimens, four types of tissue in the developing stages of the disease were analyzed: normal, benign, precancerous lesion, and tumor. Specimens were stained with HE and subjected to IHC. Normal epithelium stained less for $\alpha \mathrm{v}\beta 6$ than it did for EGFR. In addition, the soft tissue and stroma stained negative for both, allowing clear discrimination of the soft tissue margins [Fig. 1(a)].

Fig. 1

$\alpha \mathrm{v}\beta 6$ is overexpressed in the developing stages of head and neck squamous cell carcinoma. (a) Paraffin sections of human tissues ($n=6$) were subjected to HE staining and IHC staining for $\alpha \mathrm{v}\beta 6$ and EGFR (shooting magnification, $20\times$, $63\times$). (b) Quantitative analysis of $\alpha \mathrm{v}\beta 6$- and EGFR-positive areas. (c) WB detection of $\alpha \mathrm{v}\beta 6$ and EGFR in tumor tissues with different degrees of malignancy. $\alpha \mathrm{v}\beta 6$ was expressed in some benign tissues. However, the expression level was lower than that in tumor tissues. (d) Quantitative analysis of $\alpha \mathrm{v}\beta 6$ and EGFR WB assays ($n=6$, not significant [ns]: $p>0.05$, ***: $p<0.001$, **: $p<0.01$). (e) Reverse transcription-quantitative PCR detection of $\alpha \mathrm{v}\beta 6$ and EGFR in tumor tissues with different degrees of malignancy ($n=6$, ns: $p>0.05$, *: $p<0.05$, ***: $p<0.001$).

Quantification, measured as the $\alpha \mathrm{v}\beta 6$/EGFR-positive area (protocol described in Sec. 2), revealed a mean value for the two types ($\alpha \mathrm{v}\beta 6$ versus EGFR) as follows: normal tissue, $4.24\pm 0.40\%$ and $8.25\pm 0.75\%$; tumor tissue, $28.58\pm 4.18\%$ and $10.25\pm 1.23\%$ ($p<0.0001$ and $p<0.05$, respectively) [Fig. 1(b)]. The positive area was proportional to the malignancy of the tumor. The levels of $\alpha \mathrm{v}\beta 6$ were also examined using WB analysis [Fig. 1(c)].

The quantification methods included PCR-based assays and WB. The result revealed that the expression level of $\alpha \mathrm{v}\beta 6$ was remarkably upregulated in HNSCC samples compared to that with normal tissues. The differences and amounts of $\alpha \mathrm{v}\beta 6$ expression were higher than those of EGFR pre- and post-transcription. For micro-RNA expression analysis, the $\alpha \mathrm{v}\beta 6$ group was as follows: $\text{normal}=1\pm 0.09$; $\text{benign}=2.31\pm 0.24$; $\text{precancerous}=2.74\pm 0.09$; and $\text{tumor}=5.14\pm 0.38$. The EGFR group was as follows: $\mathrm{normal}=1\pm 0.08$; $\text{benign}=1.12\pm 0.12$; $\text{precancerous}=1.15\pm 0.11$; and $\text{tumor}=1.22\pm 0.10$. For WB analysis, the $\alpha \mathrm{v}\beta 6$ group was as follows: $\text{normal}=0.16\pm 0.02$; $\text{benign}=0.33\pm 0.04$; $\text{precancerous}=0.52\pm 0.06$; and $\text{tumor}=0.55\pm 0.06$. The EGFR group was as follows: $\text{normal}=0.45\pm 0.03$; $\text{benign}=0.63\pm 0.04$; $\text{precancerous}=0.44\pm 0.04$; $\text{tumor}=0.66\pm 0.05$ [Figs. 1(d) and 1(e)].

3.2.

Characterization of the $\alpha \mathrm{v}\beta 6$-Targeted Imaging Probe

The peptide selected for probe targeting was delivered using the A20FMDV2 sequence (NAVPNLRGDLQVLAQKVART). ICG was added to the N-terminus of the polypeptide [Fig. 2(a)]. The probe was also analyzed using high-performance liquid chromatography. Therefore, the potential toxicity of the probe is a major concern. Using the CCK8 assay, we found that ICG and the ICG-$\alpha \mathrm{v}\beta 6$ peptide probe had extremely low toxicity. Toxicity mainly depended on the concentration of the probe, as shown in Figs. 2(b) and 2(c). The results showed that a concentration $<1\mu \mathrm{M}$ did not influence cell growth. The HIOEC cell line was used as a control [Figs. 2(b) and 2(c)].

Fig. 2

Physicochemical properties and cytotoxicity of the ICG-$\alpha \mathrm{v}\beta 6$ peptide probe. (a) Sketch of the molecular structure of the ICG-$\alpha \mathrm{v}\beta 6$ peptide probe. (b) CCK8 toxicity assay in different tongue cancer cell lines with different concentrations of the N-succinimidyl ester (NHS)-ICG fluorescent probe ($n=3$, set probe concentration gradients: 0.10 nM, 100 nM, $1\mu \mathrm{M}$, $10\mu \mathrm{M}$, $100\mu \mathrm{M}$, and $300\mu \mathrm{M}$). (c) CCK8 toxicity assay for different concentrations of the ICG-$\alpha \mathrm{v}\beta 6$ peptide probe in different tongue cancer cell lines ($n=3$, set probe concentration gradients: 0.10 nM, 100 nM, $1\mu \mathrm{M}$, $10\mu \mathrm{M}$, $100\mu \mathrm{M}$, and $300\mu \mathrm{M}$).

3.3.

Validation of Probes for In Vitro Imaging

The ability of the probe to identify high and low $\alpha \mathrm{v}\beta 6$-expressing tumors was assessed in three kinds of HNSCC cells (CAL27, SCC9, and FaDu) using HIOEC cells as the control.

$\alpha \mathrm{v}\beta 6$ and EGFR levels in the cells were verified using WB [Fig. 3(a)]. Densitometric quantification of the blots showed $\mathrm{FaDu}>\mathrm{CAL}27>\mathrm{SCC}9>\text{control}$ [Fig. 3(b)]. Densitometric quantitation of the quantitative polymerase chain reaction (PCR) showed that $\mathrm{FaDu}=\mathrm{CAL}27>\mathrm{SCC}9>\text{control}$ [Fig. 3(c)]. To determine the selective cell binding and uptake of the probe, cells were incubated with it for 30 min. This was corroborated using confocal microscopy, where the cellular accumulation of ICG and the ICG-targeting $\alpha \mathrm{v}\beta 6$ fluorescent imaging agent could be clearly seen [Fig. 3(d)], and fluorescence intensity quantification showed that $\mathrm{HIOEC}<\mathrm{SCC}9<\mathrm{CAL}27<\mathrm{FaDu}$ [Fig. 3(e)].

Fig. 3

Validation of probes for in vitro imaging. (a) WB detection of $\alpha \mathrm{v}\beta 6$ and EGFR in different cell lines. (b) Quantitative analysis of $\alpha \mathrm{v}\beta 6$ and EGFR WB assays in different cell lines ($n=3$, ***: $p<0.001$). (c) Reverse transcription-quantitative PCR detection of $\alpha \mathrm{v}\beta 6$ and EGFR in different cell lines ($n=3$, *: $p<0.05$, **: $p<0.01$, and ***: $p<0.001$). (d) ICG and the ICG-$\alpha \mathrm{v}\beta 6$ peptide probe were incubated in different cell lines for laser confocal microscopy imaging (probe concentration set to $1\mu \mathrm{M}$, microscopy magnification $10\times$). (e) Fluorescence intensity statistics of different cell lines. (f) Laser confocal microscopy imaging of CAL27 incubated with different concentrations of N-succinimidyl ester (NHS)-ICG and ICG-$\alpha \mathrm{v}\beta 6$ peptide probe (set probe concentration gradients: 30 nM, 100 nM, 300 nM, $1\mu \mathrm{M}$, $3\mu \mathrm{M}$; microscopy magnification $40\times$). (g) Fluorescence intensity statistics of different probe concentrations of CAL27 (not significant [ns]: $p>0.05$, *: $p<0.05$, and **: $p<0.01$).

The ICG-targeting $\alpha \mathrm{v}\beta 6$ fluorescent agent showed a better imaging effect than did ICG. The quantitative metric was “fluorescence intensity,” as shown in Fig. 3(g). Based on the growth rate and fluorescence intensity, CAL27 was selected as the experimental model to identify the suitable concentration for fluorescence imaging. Optimal probe concentration was determined by evaluating different concentrations: 30 nM, 100 nM, 300 nM, $1\mu \mathrm{M}$, and $3\mu \mathrm{M}$ [Fig. 3(f)]. The mean fluorescence intensity quantification showed an increasing trend below $1\mu \mathrm{M}$. However, there was no difference in imaging when using 3 or $1\mu \mathrm{M}$ [Fig. 3(g)].

3.4.

ICG Targeting $\alpha \mathrm{v}\beta 6$ Fluorescent Probe Accumulates in Tumor Tissues and Identifies Margins

The animals were divided into three groups: tumor, inflammation, and normal ($n=6$ each). Two animal models were created for this experiment.

Tumor models were established on the tongues of NSG mice. Tumor growth was evident after an incubation period of 7 days for the CAL27 xenografts and was recorded and observed after injection. After the tumor diameter had increased to nearly $50{\mathrm{mm}}^3$, the mice were euthanized, and fluorescence imaging was subsequently performed. An inflammation model was established concomitantly with the tumor group. Mice were imaged 24 h after injection with the $\alpha \mathrm{v}\beta 6$-targeted fluorescent imaging agent. The probe was administered intravenously, and tumors were imaged using an IVIS epifluorescence-injected imaging system. We identified the $\alpha \mathrm{v}\beta 6$-targeted imaging probe in the areas of the tongue where tumors were present, whereas a lower signal was observed in the surrounding normal tissues and control tongues [Fig. 4(a)].

Fig. 4

In vivo imaging and histological analysis with fluorescent probes. (a) IVIS assay of the animal model through vein injection of the fluorescent probe ($n=6$, injection concentration: $100\mu \mathrm{M}$, injection dose: $100\mu \mathrm{L}$). (b) Relative $\alpha \mathrm{v}\beta 6$ light intensity ($\mathrm{p}/\mathrm{s}/{\mathrm{cm}}^2/\mathrm{sr}\mu \mathrm{w}/{\mathrm{cm}}^2$) (ns, not significant). (c) Frozen sections of probe-labeled tumor tissue were examined using IF ($n=6$, section thickness $5\mu \mathrm{m}$, microscopy magnification $20\times$, $63\times$). (d) IF detection of the probe in frozen sections of inflammation tissue ($n=6$, section thickness $5\mu \mathrm{m}$, microscopy magnification $20\times$, $63\times$). (e) SBRs of the ICG-$\alpha \mathrm{v}\beta 6$ peptide probe. Receiver operating characteristic curves for the whole tumor. (f) In vivo labeling of frozen sections of tumor tissue for HE, IHC, and IF detection (section thickness $5\mu \mathrm{m}$, microscopic magnification $2\times$, $63\times$).

In the $\alpha \mathrm{v}\beta 6$-targeted probe imaging group, the average fluorescence signal in the tumor area was 12 times higher than that in the other two groups. The signals were similar between the inflammatory and normal groups.

It is clinically difficult to distinguish between inflammation and tumors with the naked eye; therefore, we established an inflammation model to verify the probe’s ability to distinguish between the two different types. Control and inflamed tongues had significantly lower relative radiance efficiency signals ($p<0.001$) than xenografted tongues [Fig. 4(b)].

3.5.

$\alpha \mathrm{v}\beta 6$-Targeted Imaging Probe Accumulation in Margins and Sensitivity/Specificity Analysis

We analyzed the in situ diagnostic performance of $\alpha \mathrm{v}\beta 6$-targeted fluorescent imaging agent tumor uptake using the immunofluorescence (IF) results of the slices. We observed that the tumor area had higher fluorescence intensity than the normal area ($20\times$ and $63\times$ magnification) in the $\alpha \mathrm{v}\beta 6$-ICG group. However, the ICG group showed little difference in fluorescence intensity between the tumor area and the tissue adjacent to cancer cells [Fig. 4(c)].

The signal-to-background ratio (SBR) was obtained by dividing the fluorescence intensity of the selected region by the average intensity of the normal tissue region. Inflammatory tissue fluorescence intensity was low and showed little difference between the two probes [Fig. 4(d)].

The SBR was $2.94\pm 1.13$ standard deviation in the tumor group and $0.97\pm 0.03$ standard deviation in the inflammation group [Fig. 4(e)].

A receiver operating characteristic curve was constructed, which demonstrated an area under the curve of 0.965 with a SBR of 1.52, corresponding to a sensitivity of 91.67% and specificity of 88.33% [Fig. 4(f)].

The $\alpha \mathrm{v}\beta 6$-targeted peptide probe was most intense at the periphery of tumors. The dotted line divides the tumor into three parts: normal, border, and tumor. Confocal microscopy images of tumor samples showed abundant probes in $\alpha \mathrm{v}\beta 6$-expressing tumor cells. The border of the tumor had higher fluorescence intensity than the interior of the tumor. In benign samples, $\alpha \mathrm{v}\beta 6$ staining was confined to the thin $\alpha \mathrm{v}\beta 6$-expressing basal layer of the epithelium [Fig. 4(f)].