Open Access
Add to BibSonomyAbstract
The purpose of this study was to develop a more powerful blood analysis method based on polarized surface enhanced Raman spectroscopy (SERS) technology for non-invasive and sensitive colorectal cancer (CRC) detection. The efficiency of different polarized scattering signals (non-polarization, parallel polarization and perpendicular polarization) on blood serum SERS was explored for the first time. Results demonstrated that polarized SERS was more sensitive to explore distinctive spectral differences between cancer and normal groups. And higher diagnostic accuracy of 91.6% could be achieved using polarized SERS integrated with PCA-LDA for classification of the two serum groups in comparison to conventional SERS technology. This exploratory study demonstrated that the nanobiosensor based on polarized SERS technique in conjunction with PCA-LDA provided a novel strategy for blood SERS analysis, and had the potential as a clinical complement for CRC screening.
© 2016 Optical Society of America
1. Introduction
Colorectal cancer (CRC) is the third leading cause of cancer death worldwide with approximately 1.2 million new cases diagnosed and a mortality of over 0.6 million each year, accounting for almost 9% of all cancer deaths [1]. CRC usually requires intensive treatment, which makes a high risk of complication and high costs to patients. Despite great improvement in its treatment, almost half of CRC patients eventually die of metastatic disease [2]. Recent report from Colorectal Cancer Statistics [3] shows that 5-year relative survival rate is 90.3% for CRC tumors diagnosed at a localized stage, however that will tremendously decline to 70.4% and 12.5% for regional and distant stages in CRC, respectively. Hence, early screening coupled with proper treatment is of paramount importance to reduce the risk of CRC development and improve the survival rate. Additionally, the progression of CRC from adenomatous polyps to invasive cancer always takes a long and asymptomatic period, which makes it more suitable for population screening than other types of cancer [1]. As a noninvasive, cheap and convenient assay, faecal occult blood test (FOBT) is the most commonly used method for CRC population screening. However, it shows a limited sensitivity (approximately 50-75%) for adenomas. To improve the reliability, faecal immunochemical test (FIT) was developed with several advantages over FOBT, such as a higher detection sensitivity (approximately 60-85%), a higher population uptake and automated procedures. Despite these improvements, FIT still has a number of shortcomings, including less sensitive for lesions in the proximal colon and lack of test standardization of hemoglobin concentration. Flexible sigmoidoscopy (FS) is another alternative method for CRC screening with an excellent sensitivity (approximately 95% in the distal colon). Although FS can take biopsy and requires no full bowel preparation and sedation, it has the disadvantages of inability to visualize isolated proximal neoplasias. Colonoscopy followed by histopathological examinations currently remains the gold standard and is often the final assessment tool in CRC screening, though this approach depends on the experience of the physician and is invasive and impractical for mass screening of high-risk patients having multiple suspicious lesions [1, 2 ]. Therefore, it would be imperative clinical value to develop a new strategy capable of providing reliable, effective, convenient and non-invasive method for improving the screening of CRC.
Surface enhanced Raman spectroscopy (SERS) is an ultra-sensitive vibrational optical technology being capable of providing spectroscopic ‘fingerprints’ of specific biomolecular compositions and structures via inelastic light scattering [4–8 ]. Attributed to an electromagnetic enhancement and a chemical enhancement, Raman scattering signals of analytes can be dramatically increased with an enhancement factor of 104–108 over conventional Raman spectroscopy, by adsorbing analytes onto a nanoscale roughened metal surface [7, 9, 10 ]. Recently, SERS technology being capable of achieving single molecule detection has attracted considerable interest as a sensitive tool for biomedical applications, and particularly shows promise for human cancer-related detection. A number of studies based on SERS have been carried out to detect tumor biomarker such as proteins, RNA and DNA [11]. For instance, Chon et al. have developed a quick and reproducible SERS-based immunoassay assay which allows a low limit of detection of 1−10 pg/ml of well-known marker, carcinoembryonic antigen (CEA), to be detected for lung cancer diagnostic application [12]. Additionally, Hu et al. demonstrated the ability to detect the methylation status in the tumor suppressor gene p16 using a novel DNA methylation assay based on single base extension reaction and SERS method [13]. Although polymerase chain reaction (PCR) is often used for identification of these biomarkers, it has the disadvantages of expensive reagents and complex sample preparation. Besides, SERS holds many advantages over fluorescence method including narrow spectral width of Raman peaks (typically 10–100 times narrower than fluorescence peaks), minimal photobleaching and multiplexing capabilities under a single excitation light [4, 7, 11 ]. In addition, important advances have also been achieved in the development of in vivo SERS imaging in cells and tissues for cancer detection, and the potential utility of that is highly valuable in the future [14]. Currently, SERS technology has shown a great potential as a complement tool for cancer detection.
Human blood is an ideal biological sample for non-invasive cancer diagnosis compared with other samples such as cell and tissue, as blood can be collected conveniently and contains rich biomarkers providing a comprehensive overview of body physiology and disease status. Recently, blood-SERS has been shown to be an attractive analysis method for detection of tumor biomarker and circulating tumor cells [15, 16 ]. In particular, our group was the first to apply SERS to blood investigation with the aim to develop a label-free nanobiosenor for non-invasive cancer detection [10, 17–20 ]. A further work by us has been done on exploring the effect of different laser excitation polarizations on blood-SERS analysis. Interesting results showed that the diagnostic performance depended on the polarization status of laser excitation, and the best detection result could be achieved by left-handed circularly polarized laser excitation [21]. However, the potential of scattering polarization based blood-SERS method has not been assessed in detail for cancer detection.
Thus, the primary aim of this work was to apply polarized SERS technology to probe biochemical changes in blood serum samples between normal and CRC groups. The effects of different polarized scattering signals (non-polarization, parallel polarization and perpendicular polarization) on blood-SERS test were evaluated for improving non-invasive CRC detection. Multivariate statistical methods, including principal component analysis combined with linear discriminant analysis (PCA-LDA), were employed to differentiate the blood serum SERS spectra obtained from the two groups. This work may further promote the label-free blood SERS technique into practical clinical cancer detection.
2. Materials and methods
2.1 Blood serum samples collection
Human blood experiments were performed in accordance with relevant guidelines and regulations, and approved by the ethical committee at our institution (Fujian Provincial Cancer Hospital, Fujian, China). In addition, the informed consents were obtained from all subjects. We conducted a study of cancer and normal blood serum samples obtained from 38 colorectal cancer patients and 45 healthy volunteers at Fujian Provincial Cancer Hospital, Fujian, China. The diagnoses for the cancer patients were confirmed by histopathological tests. The mean age for the cancer group was 55 years and for the control group was 53 years. Before SERS measurement, a 10 μl blood serum was mixed with a 10 μl prepared silver colloid to create a homogenous mixture. The mixture was incubated at 4 °C for 2 h. Finally, a drop of this mixture was transferred onto an aluminum plate for SERS measurement.
2.2 Silver colloidal solution preparation
Stable Ag colloidal solutions were produced following the reported protocol [22]. In brief, a total of 90 mg silver nitrate (Shenyang gaohong chemical co., LTD) was dissolved in distilled water (500 ml, 45 °C) with a rolling boil under stirring. When the AgNO3 solution started boiling, a 1.0% solution of sodium citrate (10 ml) was dropped into that, while boiling and stirring process was continued. After the mixture had boiled for 90 min, the heating mantle was removed and the solution was cooled down naturally. The final solution was analyzed using UV-vis spectroscopy, and represented an absorption peak at 426 nm. Additionally, prepared Ag nanoparticles (NPs) were characterized by transmission electron microscopy (TEM), following a normal distribution with a mean diameter of 60 nm and standard deviation of 3 nm (Fig. 1 ). Finally, the Ag colloidal solution was concentrated by centrifugation at 10000 rpm for 10 min, and the final concentration was obtained for later use. When the serum samples were mixed with the prepared Ag colloids, the biomolecules contained in blood serum were nonspecifically adsorbed onto the surfaces of Ag NPs leading to an aggregation. Thus, the Raman scattering signal of serum can be extraordinarily enhanced due to the electromagnetic field enhancement based on strong localized surface plasmon resonance.
2.3 Polarized SERS measurement
The blood serum SERS spectra were detected by using a Raman micro-spectrometer (inVia System; Renishaw plc, Gloucestershire, United Kingdom) equipped with a 785 nm laser. The SERS scattering signals of serum samples in the wavelength region of 400–1750 cm−1 were recorded within 10 s integration time with 20 × objective, using Peltier CCD detector. The software package WIRE 2.0 (Renishaw) was used for spectral acquisition and analysis.
In order to compare linearly polarized (i.e., perpendicular and parallel polarized) SERS spectra of blood samples with the nonpolarized SERS spectra of the same samples, a linear polarizer was placed in the front of CCD detector. The parallel polarized and perpendicular polarized SERS measurement can be performed by changing the direction of the inserted polarizer to be parallel or perpendicular with respect to the polarization direction of the excitation laser source. The nonpolarized SERS spectra were detected without the inserted polarizer. The scattered signal was sequentially detected as non-, perpendicular-, and parallel-polarizations. Figure 1 represents a sketch of the polarized SERS measurement used in this study.
2.4 Spectral data analysis
Before multivariate analysis, a preprocessing algorithm (Vancouver Raman Algorithm) based on fifth-order polynomial fitting was performed for the raw spectral data in order to remove the prominent auto-fluorescence background contained in the acquired SERS data and yield the blood serum SERS signals alone. Each of background-subtracted SERS spectrum was then normalized to the area under the curve in the range of 400–1750 cm−1 to minimize the spectral intensity variability for enabling a better comparison of the spectral characteristics (spectral shapes or peak intensities) among different samples for further spectral analysis. Mean spectral intensities of eight SERS peaks at around 494, 591, 638, 813, 888, 1135, 1578 and 1655 cm−1 were specifically compared between normal and cancer groups under the three polarized detection states. T-test (pairwise comparison of blood groups) was further used for assessing diagnostic utility of these peaks for serum samples classification. Error bars (whiskers) represent the 1.5-fold inter-quartile range.
The multivariate analysis method used for the blood SERS spectra studies has been described in detail in previous report [10]. Briefly, this method represents a combination of principal component analysis (PCA) and linear discriminant analysis (LDA), achieving an efficient classification of SERS data belong to normal and cancer blood samples. In present work, PCA was first performed on the normalized data set to identify a new set of orthogonal variables, called principal components (PC), which is capable of reflecting the differences between different groups. Independent-sample T test was then used to explore diagnostically significant PC scores (PCs) for each case using an alpha of 5%. Finally, significant PCs (p<0.05) were selected as the input for LDA model for correctly predicting the blood samples (i.e. normal vs. cancer). Herein, leave-one spectrum-out and cross-validation methods were carried out in LDA. To compare the performances of different polarized SERS for discriminating between normal and cancer groups using the same blood samples, receiver operating characteristic (ROC) curves were utilized by changing the thresholds to determine discrimination sensitivity and specificity for all samples. Algorithms for above-mentioned analysis were implemented using the SPSS software package (SPSS Inc., Chicago).
3. Results
To assess the feasibility of polarized SERS technology for biomolecular analysis on blood serum samples for CRC detection, high quality SERS spectra of normal and cancer samples were first obtained and compared to explore the changes of biomolecules in serum associated with cancer transformation. Figures 2(a)–2(c) showed the comparison of normalized mean SERS spectra between normal (n = 45) and CRC (n = 38) serum samples obtained under non-polarized, parallel polarized and perpendicular polarized SERS, respectively. The standard deviations (SD) for each group were also calculated to understand spectral reproducibility within the group, overlying as shaded color fill. Prominent SERS peaks located at around 494, 591, 638, 725, 813, 888, 1003, 1072, 1135, 1206, 1332, 1578 and 1655 cm−1, which could be attributed to the biochemical bonds of ring vibration of L-arginine, amide-VI, C-C twisting mode of tyrosine, hypoxanthine, C-C stretching mode of collagen, C-O-H bending mode of D-galactosamine, C-C symmetric stretch ring breathing of phenylalanine, C-C stretching of lipid, D-mannos, C-C6H5 stretching of tryptophan, tryptophan, phenylalanine and C = O stretching of amide-I (Table 1 ) [10, 15, 19, 20, 23 ], respectively, were clearly observed in both normal and cancer blood serum samples. Although there was a close correspondence between the patterns of spectral shapes and frequencies in these two groups (normal vs. cancer), small spectral relative intensities changes were evident for some vibration peaks when normal (blue lines) and cancer (red lines) serum SERS spectra were compared in great detail. These spectral differences could be viewed more clearly in the difference spectrum shown on the bottom of each figure in Figs. 2(a)-2(c). For example, cancer serum showed higher intensities at 725, 1003 and 1332 cm−1, while lower at 1072 cm−1, compared with normal sample, which could be consistently obtained under all polarized detection states (non-polarization, parallel polarization and perpendicular polarization).
Fig. 2 (a-c) Comparison of normalized mean SERS spectra from 45 normal and 38 CRC blood serum samples under different polarized SERS. The shaded areas represent the standard deviations of the means. Also shown at the bottom is the difference spectrum. (d) Different polarized SERS spectra from a same serum sample.
Table 1. The peak positions and tentative assignment of major vibrational bands observed in serum samples.
Besides, other eight SERS peaks at around 494, 591, 638, 813, 888, 1135, 1578 and 1655 cm−1 were specifically identified, as the intensities of these peaks exhibited different changes between normal and cancer groups under the three polarized detection states. Comparison of the intensities with ± SD at each of the eight SERS peaks was illustrated in the bar-chart (Fig. 3 ). It can be seen that cancer group showed lower intensities at 494, 813, 888, 1135 and 1655 cm−1 as compared to normal group under all polarized detection states with different levels. However, the intensities of the peaks at 591 and 638 cm−1 were found to be slightly higher in cancer group under nonpolarized state, whereas the two peaks showed lower in cancer group under parallel and perpendicular polarized states. In addition, the peak at 1578 cm−1 showed an opposite trend compared to the peaks at 591 and 638 cm−1. Moreover, T-test was employed to further identify the intensities changes of these peaks. Different degrees of diagnostic utility for discriminating the two groups (normal vs. cancer) based on the same peak under different polarized detection states were presented in Fig. 3. Results indicated that most of the SERS peaks except for 1578 and 1655 cm−1 showed greater efficacy in classification of the normal and cancer groups under lineally polarized states in comparison to nonpolarized state. These observations indicated the diagnostic potential of polarized SERS technology for CRC detection based on blood serum analysis.
Fig. 3 Histogram of the eight special SERS peak intensities for the two blood sample types under different polarized SERS: (a) 494 cm−1, (b) 591 cm−1, (c) 638 cm−1, (d) 813 cm−1, (e) 888 cm−1, (f) 1135 cm−1, (g) 1578 cm−1 and (h) 1655 cm−1. Error bars (whiskers) represent the 1.5-fold interquartile range. *p <0.05 (pairwise comparison of blood groups with T-test).
The multivariate statistical method based on PCA-LDA by incorporating the whole SERS spectrum was employed to further explore the most diagnostically significant SERS features for improving serum samples classification. T-test on the PCs obtained by PCA showed that three PCs (PC1, PC2 and PC3) accounting for 66.93% of the variance were diagnostically significant (p <0.05) for discriminating cancer samples from normal samples, under non-polarized state (Fig. 4(a) ). Similarly, PC1, 3 and 6, and PC1, 5 and 6 were found to be the most diagnostically significant for discriminating the two groups respectively, under parallel and perpendicular polarized states. Some PC characteristics (Figs. 4(a), 4(c) and 4(e)), such as spectral peaks and shapes were similar to those of blood SERS spectra previously showed in Fig. 2. Besides, Figs. 4(b), 4(d) and 4(f) showed classification models of normal and cancer groups generated by the correlations between the optimal PCs under different polarized SERS, and different samples could be clustered into two separate groups on the basis of different combinations of PCs. For example, the corresponding separation lines classified cancer from normal groups with a sensitivity of 78.9% (30/38) and 63.2% (24/38); a specificity of 64.4% (29/45) and 80% (36/45) under nonpolarized SERS (Fig. 4(b)). Using parallel polarized SERS, the sensitivity of 76.3% (29/38) and 86.8% (33/38); the specificity of 88.9% (40/45) and 80% (36/45) could be achieved (Fig. 4(d)). Meanwhile, using perpendicular polarized SERS, the sensitivity of 71.1% (27/38) and 89.5% (34/38); the specificity of 86.7% (39/45) and 77.8% (35/45) were obtained (Fig. 4(f)). Results showed that different combinations of significant PCs would provide different levels of accuracy for samples classification under a certain polarized SERS.
Fig. 4 (a, c, e) The three diagnostically significant principal components (PCs) calculated from SERS spectra, revealing the diagnostically significant spectral features for serum classification under different polarized SERS. (b, d, f) Scatter plots of the corresponding PCs for normal and cancer groups under different polarized SERS.
These PCs selected in each of polarized SERS were subsequently loaded into the LDA respectively, in which the classification criterion was identified using the scatter measure of within-class and between-class variance, to increase the efficiency of blood samples classification. Figures 5(a)-5(c) show the posterior probabilities belonging to the normal and cancer groups as calculated from different polarized SERS in the LDA model. Using a discrimination threshold of 0.4798, 0.6620 and 0.2514, the diagnostic sensitivity for detecting CRC cancer was 81.6%, 89.5% and 100% for non-polarized, parallel polarized and perpendicular polarized SERS, respectively. The corresponding diagnostic specificities for each method were 75.6%, 93.3% and 84.4%, respectively. These classification results are summarized in Table 2 .
Fig. 5 (a-c) Scatter plots of the posterior probabilities belonging to the normal and CRC groups using the PCA-LDA together with leave-one-out and cross-validation method. The separate line yields a diagnostic accuracy of 78.3%, 91.6% and 91.6% for nonpolarized, parallel and perpendicular polarized blood SERS spectral classification, respectively. (d) Comparison of ROC curves of the discrimination results for SERS spectra with different polarized SERS. The integration areas under the ROC curves (AUC) are 0.905, 0.962 and 0.964, respectively, for the three polarized SERS, respectively.
Table 2. Results of polarized SERS spectra prediction for differentiating cancer from normal blood samples using PCA-LDA diagnostic algorithma.
To evaluate and compare the performance of the PCA-LDA based diagnostic algorithms derived from different polarized SERS methods, the ROC curves (Fig. 5(d)) were generated from the scatter plots in Figs. 5(a)-5(c) at different threshold levels. The integration areas under the ROC curves were 0.905, 0.962 and 0.964 for non-polarized, parallel and perpendicular polarized SERS, respectively. Results demonstrated that linearly polarized SERS yielded a better diagnostic capability than non-polarized SERS for CRC serum detection, as illustrated by the improvement in the diagnostic sensitivity and specificity.
4. Discussion
Human blood has attracted more attention for cancer diagnosis, because the content of biomolecules, specifically proteins and DNA/RNA contained in blood will alter due to apoptosis and necrosis of cells with tumor formation. However, there are multiple molecules are present in one sample and the changes of these biomolecules are extremely small, which make it a great challenge to achieve efficient blood detection. SERS technology is capable of single molecule detection and multiplexed analysis, and can therefore allow for rapid and sensitive detection of certain biomarkers in human blood.
Distinctive SERS spectral differences between cancer and normal groups reflected molecular changes in quantity or structure associated with CRC formation. For instance, the relative peak intensities at 1003 cm−1 and 1332 cm−1 attributed to phenylalanine and tryptophan, respectively, showed higher percentage signals for cancer serum compared to the normal, indicating the elevated concentration of phenylalanine and tryptophan relative to the total Raman-active components in cancer samples. Huang et al. also observed an increase of phenylalanine and tryptophan in lung cancer using Raman spectroscopy [24]. Similar change of tryptophan residues was presented in skin tumor [25]. In addition, there was a relative increase of hypoxanthine band (725 cm−1) in intensity for cancer group, suggesting that CRC serum may be associated with an increase in the relative amount of DNA/RNA bases-related substances. Hypoxanthine is a ubiquitous metabolite which is produced in the highly conserved purine degradation pathway in both prokaryotic and eukaryotic cells [15]. The significant change of hypoxanthine and related DNA/RNA bases may reveal an abnormal metabolism of DNA/RNA bases in the blood of CRC subjects, due to the probable apoptosis and necrosis, or release of intact cells in the bloodstream and their subsequent lysis [10]. Furthermore, the higher levels of cell-free circulating DNA/RNA in the plasma or serum of cancer patients have been observed by previous studies [26]. For example, Frattini et al. reported a quantification detection of blood cell-free DNA using the Dipstick Kit method. Results showed the cell-free DNA levels in plasma of colorectal cancer patients were about 25 times higher than that of healthy subjects, representing a promising method for monitoring CRC and, prospectively, for identifying high-risk individuals [27]. The above changes of peak intensities at 725, 1003 and 1332 cm−1 were similar between normal and cancer groups under the three types of polarized SERS, however other peaks presented in Fig. 3 demonstrated distinctive changes under different polarized SERS states. It can be seen that greater efficacy in classification of the normal and cancer groups was achieved by lineally polarized SERS in comparison to non-polarized SERS. Although the Raman spectrum of different polarized SERS was similar (Fig. 2(d)), the difference spectrum intensity of polarized SERS was more drastic than that of non-polarized SERS (Fig. 2(a)-2(c)), which further confirmed that polarized SERS was capable of achieving greater efficacy in serum classification. Therefore, polarized SERS had a diagnostic potential for exploring the specific changes of biomolecules in serum for CRC detection.
It should be noted that the simplistic empirical analysis above only uses limited SERS peak features, and most of the information contained in the SERS spectra has not been utilized. Besides, as evident in Fig. 2, the SERS spectral patterns between normal and cancer groups were very similar, thus a multivariate analysis method based on PCA-LDA was employed in this work to extract all possible diagnostic information contained in serum SERS spectra for improving the diagnostic efficiency of CRC blood SERS analysis. This efficient diagnostic approach has been widely applied for Raman spectral analysis in cancer-related samples detection, including cell [28, 29 ], tissue [30] and blood [10]. As shown in Fig. 4, the combinations of significantly diagnostic PC scores were performed for classification of normal and cancer groups. The normal and cancer samples could be largely clustered into two separate groups with different degrees of overlap on the basis of different combinations of PCs and different polarized SERS states. In addition, the regional distribution of cancer samples was more dispersed than that of normal group, presented in Fig. 4. LDA was then used in conjunction with PCA to increase the efficiency of serum samples classification. The diagnostic accuracy of 91.6% could be achieved for differentiating cancer from normal serum using linearly polarized SERS (parallel and perpendicular polarization), which had almost a 13% improvement in diagnostic accuracy compared with the non-polarized SERS (Table 2). Compared with our previous report using SERS based on Au NPs on blood analysis for colorectal cancer [18], non-polarized SERS in this work has lower diagnostic accuracy. However, the diagnostic accuracy is close to previous result using polarized SERS. ROC analysis (Fig. 4(d)) further confirms that blood polarized SERS together with PCA-LDA diagnostic algorithm is more effective for classification between cancer and normal serum groups in comparison to conventional SERS measurement. This may be attributed to the unique ability of polarized SERS for biomolecular orientation detection. Since the Raman scattered light will shift to another frequency correlated with specific molecular composition and structure respect to the incident excitation light, the polarization of Raman signals may reflect the subtle changes of biomolecular orientation, which makes it possible to differentiate preferentially oriented chemical bonds from randomly oriented chemical bonds within the biological sample [31]. There are inevitable changes of biomolecular orientation in blood sample associated with cancer transformation, and that will be sensitively detected by polarized SERS. The other probable reason for the increased diagnostic efficiency with polarized SERS is that most biomolecules are chirality, which makes the changes of biomolecules more sensitive to the polarized SERS [21]. Meanwhile, we recognize that it is still a great challenge to fully understand the accurate mechanisms for the superior diagnostic performance of polarized SERS so far. The suggested approach is mainly focused on the methodology development and a pilot application for CRC detection. Further investigations and a better understanding of this novel spectroscopic approach are required.
5. Conclusion
A new blood analysis strategy based on polarized SERS was developed for CRC detection. The proposed method was capable of detecting biomolecular differences between CRC and normal serum samples, and higher diagnostic accuracy (approximately 90%) could be achieved using this method with PCA-LDA diagnostic algorithm for differentiating cancer from normal groups in comparison to conventionally non-polarized SERS. Particularly, this polarized SERS method provided a unique opportunity to detect the orientations of biomolecular chemical bonds. We believe this preliminary work will promote further study on blood SERS analysis and the results demonstrate the potential of polarized SERS technique to be a clinical complement for CRC screening.
Acknowledgments
This work was supported by the National Natural Science Foundation of China (Nos. 61575043, 61178090, 61405036 and 81101110), the major projects of international cooperation (No. 61210016), the program for Changjiang Scholars and Innovative Research Team in University (No. IRT15R10), the Science and Technology Project of Fujian Province (No.WKJ-FJ-01), Fujian province health commission Young and middle-aged talent training project (2014-ZQN-JC-6) and Fujian Provincial Natural Science Foundation Project (2015J01436).
References and links
1. S. C. Ng and S. H. Wong, “Colorectal cancer screening in Asia,” Br. Med. Bull. 105(1), 29–42 (2013). [CrossRef] [PubMed]
2. E. J. Kuipers, T. Rösch, and M. Bretthauer, “Colorectal cancer screening-optimizing current strategies and new directions,” Nat. Rev. Clin. Oncol. 10(3), 130–142 (2013). [CrossRef] [PubMed]
3. R. Siegel, C. Desantis, and A. Jemal, “Colorectal cancer statistics, 2014,” CA Cancer J. Clin. 64(2), 104–117 (2014). [CrossRef] [PubMed]
4. U. S. Dinish, G. Balasundaram, Y.-T. Chang, and M. Olivo, “Actively targeted in vivo multiplex detection of intrinsic cancer biomarkers using biocompatible SERS nanotags,” Sci. Rep. 4, 4075 (2014). [CrossRef] [PubMed]
5. E.-J. Chae, J.-H. Lee, B.-K. Oh, and J.-W. Choi, “Label-free nanobiosensor to detect infectious bacterica based on SERS,” J. Biomed. Nanotechnol. 9(4), 659–663 (2013). [CrossRef] [PubMed]
6. S.-C. Luo, K. Sivashanmugan, J.-D. Liao, C.-K. Yao, and H.-C. Peng, “Nanofabricated SERS-active substrates for single-molecule to virus detection in vitro: A review,” Biosens. Bioelectron. 61, 232–240 (2014). [CrossRef] [PubMed]
7. K. C. Bantz, A. F. Meyer, N. J. Wittenberg, H. Im, O. Kurtuluş, S. H. Lee, N. C. Lindquist, S.-H. Oh, and C. L. Haynes, “Recent progress in SERS biosensing,” Phys. Chem. Chem. Phys. 13(24), 11551–11567 (2011). [CrossRef] [PubMed]
8. X. Li, T. Yang, S. Li, L. Jin, D. Wang, D. Guan, and J. Ding, “Noninvasive liver diseases detection based on serum surface enhanced Raman spectroscopy and statistical analysis,” Opt. Express 23(14), 18361–18372 (2015). [CrossRef] [PubMed]
9. Q. Tu and C. Chang, “Diagnostic applications of Raman spectroscopy,” Nanomedicine (Lond.) 8(5), 545–558 (2012). [PubMed]
10. D. Lin, J. Pan, H. Huang, G. Chen, S. Qiu, H. Shi, W. Chen, Y. Yu, S. Feng, and R. Chen, “Label-free blood plasma test based on surface-enhanced Raman scattering for tumor stages detection in nasopharyngeal cancer,” Sci. Rep. 4, 4751 (2014). [PubMed]
11. J. Srajer, A. Schwaighofer, and C. Nowak, “Surface-enhanced Raman spectroscopy for biomedical diagnostics and imaging,” Biomed. Spectrosc. Imaging 2, 51–71 (2013).
12. H. Chon, S. Lee, S. W. Son, C. H. Oh, and J. Choo, “Highly sensitive immunoassay of lung cancer marker carcinoembryonic antigen using surface-enhanced Raman scattering of hollow gold nanospheres,” Anal. Chem. 81(8), 3029–3034 (2009). [CrossRef] [PubMed]
13. J. Hu and C. Y. Zhang, “Single base extension reaction-based surface enhanced Raman spectroscopy for DNA methylation assay,” Biosens. Bioelectron. 31(1), 451–457 (2012). [CrossRef] [PubMed]
14. M. Vendrell, K. K. Maiti, K. Dhaliwal, and Y.-T. Chang, “Surface-enhanced Raman scattering in cancer detection and imaging,” Trends Biotechnol. 31(4), 249–257 (2013). [CrossRef] [PubMed]
15. W. R. Premasiri, J. C. Lee, and L. D. Ziegler, “Surface-enhanced Raman scattering of whole human blood, blood plasma, and red blood cells: cellular processes and bioanalytical sensing,” J. Phys. Chem. B 116(31), 9376–9386 (2012). [CrossRef] [PubMed]
16. H. Ito, H. Inoue, K. Hasegawa, Y. Hasegawa, T. Shimizu, S. Kimura, M. Onimaru, H. Ikeda, and S. E. Kudo, “Use of surface-enhanced Raman scattering for detection of cancer-related serum-constituents in gastrointestinal cancer patients,” Nanomedicine (Lond.) 10(3), 599–608 (2014). [PubMed]
17. S. Feng, D. Lin, J. Lin, B. Li, Z. Huang, G. Chen, W. Zhang, L. Wang, J. Pan, R. Chen, and H. Zeng, “Blood plasma surface-enhanced Raman spectroscopy for non-invasive optical detection of cervical cancer,” Analyst (Lond.) 138(14), 3967–3974 (2013). [CrossRef] [PubMed]
18. D. Lin, S. Feng, J. Pan, Y. Chen, J. Lin, G. Chen, S. Xie, H. Zeng, and R. Chen, “Colorectal cancer detection by gold nanoparticle based surface-enhanced Raman spectroscopy of blood serum and statistical analysis,” Opt. Express 19(14), 13565–13577 (2011). [CrossRef] [PubMed]
19. D. Lin, S. Feng, H. Huang, W. Chen, H. Shi, N. Liu, L. Chen, W. Chen, Y. Yu, and R. Chen, “Label-free detection of blood plasma using silver nanoparticle based surface-enhanced Raman spectroscopy for esophageal cancer screening,” J. Biomed. Nanotechnol. 10(3), 478–484 (2014). [CrossRef] [PubMed]
20. S. Feng, R. Chen, J. Lin, J. Pan, G. Chen, Y. Li, M. Cheng, Z. Huang, J. Chen, and H. Zeng, “Nasopharyngeal cancer detection based on blood plasma surface-enhanced Raman spectroscopy and multivariate analysis,” Biosens. Bioelectron. 25(11), 2414–2419 (2010). [CrossRef] [PubMed]
21. S. Feng, R. Chen, J. Lin, J. Pan, Y. Wu, Y. Li, J. Chen, and H. Zeng, “Gastric cancer detection based on blood plasma surface-enhanced Raman spectroscopy excited by polarized laser light,” Biosens. Bioelectron. 26(7), 3167–3174 (2011). [CrossRef] [PubMed]
22. C. Munro, W. Smith, M. Garner, J. Clarkson, and P. White, “Characterization of the surface of a citrate-reduced colloid optimized for use as a substrate for surface-enhanced resonance Raman scattering,” Langmuir 11(10), 3712–3720 (1995). [CrossRef]
23. J. L. González-Solís, J. C. Martínez-Espinosa, L. A. Torres-González, A. Aguilar-Lemarroy, L. F. Jave-Suárez, and P. Palomares-Anda, “Cervical cancer detection based on serum sample Raman spectroscopy,” Lasers Med. Sci. 29(3), 979–985 (2014). [CrossRef] [PubMed]
24. Z. Huang, A. McWilliams, H. Lui, D. I. McLean, S. Lam, and H. Zeng, “Near-infrared Raman spectroscopy for optical diagnosis of lung cancer,” Int. J. Cancer 107(6), 1047–1052 (2003). [CrossRef] [PubMed]
25. L. Brancaleon, A. J. Durkin, J. H. Tu, G. Menaker, J. D. Fallon, and N. Kollias, “In vivo fluorescence spectroscopy of nonmelanoma skin cancer,” Photochem. Photobiol. 73(2), 178–183 (2001). [CrossRef] [PubMed]
26. R. M. Bremnes, R. Sirera, and C. Camps, “Circulating tumour-derived DNA and RNA markers in blood: a tool for early detection, diagnostics, and follow-up?” Lung Cancer 49(1), 1–12 (2005). [CrossRef] [PubMed]
27. M. Frattini, G. Gallino, S. Signoroni, D. Balestra, L. Battaglia, G. Sozzi, E. Leo, S. Pilotti, and M. A. Pierotti, “Quantitative analysis of plasma DNA in colorectal cancer patients: a novel prognostic tool,” Ann. N. Y. Acad. Sci. 1075(1), 185–190 (2006). [CrossRef] [PubMed]
28. Z. A. Nima, M. Mahmood, Y. Xu, T. Mustafa, F. Watanabe, D. A. Nedosekin, M. A. Juratli, T. Fahmi, E. I. Galanzha, J. P. Nolan, A. G. Basnakian, V. P. Zharov, and A. S. Biris, “Circulating tumor cell identification by functionalized silver-gold nanorods with multicolor, super-enhanced SERS and photothermal resonances,” Sci. Rep. 4, 4752 (2014). [CrossRef] [PubMed]
29. H. Huang, D. Lin, W. Chen, Y. Yu, J. Xu, Z. Liang, X. Lin, Z. Dong, and H. Shi, “Nondestructive discrimination between normal and hematological malignancy cell lines using near-infrared Raman spectroscopy and multivariate analysis,” Laser Phys. Lett. 11(8), 085601 (2014). [CrossRef]
30. S. K. Teh, W. Zheng, K. Y. Ho, M. Teh, K. G. Yeoh, and Z. Huang, “Diagnostic potential of near-infrared Raman spectroscopy in the stomach: differentiating dysplasia from normal tissue,” Br. J. Cancer 98(2), 457–465 (2008). [CrossRef] [PubMed]
31. N. S. J. Lim, Z. Hamed, C. H. Yeow, C. Chan, and Z. Huang, “Early detection of biomolecular changes in disrupted porcine cartilage using polarized Raman spectroscopy,” J. Biomed. Opt. 16, 017003 (2011).
