Laser-induced breakdown spectroscopy (LIBS) has the potential to be used as a surgical tool for simultaneous tissue ablation and elemental analysis of the ablated tissue. LIBS may be used to distinguish melanoma lesions from the surrounding dermis based on the quantitative difference of elements within melanoma lesions. Here, we measured the elements in homogenized pellets and real tissues from excised skin samples of melanoma-implanted mice. In addition, statistical analysis of LIBS spectra using principal component analysis and linear discriminant analysis was performed. Our results showed that this method had high detection sensitivity, highlighting the potential of this tool in clinical applications.
© 2015 Optical Society of America
Laser surgery is widely used in the ablation of skin lesion, providing benefits such as precise tissue destruction, including depth control, minimal invasiveness with least collateral damage, noncontact tissue removal to avoid crushing injury, and antiseptic effects [1, 2 ]. However, surgeons usually simply predict complete removal of the lesion by visual inspection of the ablated surface during laser surgery. Since malignant lesions, including the margins, should be histologically analyzed for diagnosis and complete resection, total ablation of primary malignant lesions is challenging . Furthermore, histologic verification of the absence of marginal malignant cells is important to confirm the complete removal of the lesion. The presence of residual cancer cells after incomplete excision can cause recurrence of the disease.
The incidence of cutaneous malignant melanoma has increased dramatically in recent years. It is estimated that 76,100 cases are newly developed and 9,710 cases are death from cutaneous melanoma in United States, 2014 . Almost one-third of all patients with melanoma experience recurrence of the disease, and 21.8% of patients with recurrent melanoma experience local recurrence . In one study, two cases out of a total of 19 cases of local recurrence after primary melanoma excision were caused by incomplete excision .
To avoid recurrence caused by incomplete excision, imaging modalities, such as optical coherence tomography (OCT), have been studied. However, imaging modalities that can be applied during operation require additional processing and generally cannot be used for real-time analysis of removed tissue. Recent studies have introduced the concept of smart surgical tools using laser-induced breakdown spectroscopy (LIBS) as a real-time feedback system during laser surgery . LIBS is a type of atomic emission spectroscopy that includes ablation of the sample surface with a pulsed laser beam to generate plasma and provides real-time information of elements in ablated samples. Real-time feedback using biomarkers such as the elemental composition of the ablated tissue may facilitate identification of the depth of ablation and the excision margin by controlling the laser parameters.
Therefore, in this study, we focused on the feasibility of LIBS for the differentiation of melanoma lesions from the surrounding dermis. We conducted qualitative and quantitative elemental analysis of melanomas and the surrounding dermis using homogenized pellet samples from melanoma-implanted mice. Based on these results, we also analyzed distinguishable elements from real skin tissues of melanoma-implanted mice, reflecting the clinical situation.
2.1. Animal model and sample preparation
To induce melanoma in SKH-1 mice (male; 8–10 weeks of age; Orient-Bio, Gyeonggi-do, Korea), B16/F10 cells, a well-known murine melanoma cell line, were implanted into the mice as follows. B16/F10 cells were cultured in Dulbecco’s modified Eagle medium (DMEM; Gibco/Invitrogen, Carlsbad, CA, USA) supplemented with 10% (v/v) fetal bovine serum (FBS; HyClone, Logan, UT, USA) and 1% (v/v) penicillin/streptomycin (Gibco/Invitrogen). Cells were grown in an incubator at 37°C with a humidified atmosphere containing 5% CO2. After reaching 70% confluence, B16/F10 cells were trypsinized, centrifuged, counted, and placed in individual syringes (1 × 106 cells/100 μL of 0.9% NaCl solution). The cells were then injected in the subcutaneous layer of the back skin after anesthesia with an intraperitoneal injection of ketamine (70 mg/kg; Ketamine 50; Yuhan, Seoul, Korea) and xylazine (7 mg/kg; Rompun; Bayer Korea, Seoul, Korea). After 10–12 days, the upper epidermis of the back skin, including the melanoma lesion, was removed with the tape stripping technique to eliminate the influence of epidermal surface contamination. Skin tissues including the melanoma and dermis were prepared by excision for LIBS analysis and histology. Skin samples were fixed, embedded, sliced, and stained with hematoxylin and eosin (H&E). For pelletization, excised melanomas and dermis were dried using a freeze-dryer (FreeZone Plus 12; Labconco, Kansas City, MO, USA) for 120 h at −80°C and then ground up. All ground samples were pelletized in a mold under 1 tons of pressure for 1 min using a digital hydraulic press (PIKE 181-1110; PIKE Technologies, USA). The procedures for sample preparation are shown in Figs. 1(a)–1(c) . Five pellets each for the melanomas and dermis were used for the pellet analysis (Fig. 1(b)). Three excised tissues were tested within 2 hours without any further pretreatment for mimicking clinical situation (Fig. 1(c)).
All procedures were approved by the Animal Care and Use Committee of Gwangju Institute of Science and Technology and performed in accordance with NIH guidelines (GIST-05).
2.2. LIBS measurement
LIBS analysis of the pellet and tissues were carried out using a commercial LIBS system (Applied Spectra Inc., RT250-EC). A second harmonic Q-switched Nd:YAG laser (λ = 532 nm, τ = 5 ns, top-hat profile) was used to generate plasma, and the incident laser beam was perpendicular to the sample surface. The spot diameter and laser pulse energy in pellet and tissue experiments are presented in Table 1 . The laser pulse energy was selected for the value at which the emission signals of the major elements (C, Mg, Ca, Na, H, O, N, Cl) were clearly detected, and the pulse-to-pulse stability of the laser energy was less than 1%. The plasma emission was collected by a collection lens installed at an angle of about 30° from the incident laser beam direction and delivered to the CCD spectrometer with spectral window of 187–1045 nm at a resolution of about 0.1 nm. A schematic of the LIBS measurement set up is shown in Fig. 2 . Air (15 L/min) was continuously injected at an angle of about 40° from the sample surface in pellet experiments to minimize contamination of the focusing and collection lenses from ablation-induced particles, whereas Ar gas (15 L/min) was used in excised tissue experiments to improve signal intensity as well as lens protection.
3. Results and discussion
3.1 Analysis of elements in melanomas and dermis
There are several characteristics of samples that can cause differences in the plasma intensities of melanomas and dermis. For example, melanomas and the surrounding dermis have different absorption properties at the laser wavelength of 532 nm. Generally, normal melanocytes are located in the basal layer of epidermis and contribute to epidermal pigmentation; therefore, the absorption properties of the melanoma may be similar to those of highly pigmented epidermis. According to a previous report , the absorption coefficient (μa) of highly pigmented epidermis (16.13 cm−1) at 550 nm is higher than that of dermis (7.86 cm−1). In addition, melanoma lesions have higher dryness and rigidity than the surrounding dermis. Differences in water content and rigidity may be caused by differences in the primary constituents of the tissues. As shown in Fig. 1(d), we found that melanomas had more dense cellular components containing melanin pigment compared with normal dermis, which mainly consists of collagen bundles, other extracellular matrix components, such as glycosaminoglycan, and few fibroblasts . In the dermis, glycosaminoglycans, including hyaluronic acid, bind up to 1000 times their volume in water and play a major role in the hydration of the skin .
To examine the differences in relative intensities among constituent elements in melanomas and dermis, all spectra were normalized by the peak value of C(I) 247.856 nm emission line, which was shown to be nearly independent of malignancy status of tissue in a previous report . Additionally, the first shot was excluded from analysis to eliminate possible errors caused by sample contamination during handling. The LIBS spectra from the second to the fifteenth shots of the same spots in pellet and excised tissue were averaged. The normalized spectra from all shots in the melanoma and surrounding dermis are shown in Fig. 3 . The emission peaks of known components of biological tissues, i.e., C, Mg, Ca, Na, H, Na, K, O, and Cl, were identified in the spectra from pellet and tissue samples in the range from 200 to 900 nm. From Fig. 3, in the spectra from pellet and tissue samples, it is observed that emission signals of these known elements are detected from both melanoma and dermis, but the intensities of these elements are different.
The normalized peak intensities of Mg and Ca in the melanoma were significantly higher than those in the dermis for both pellet and tissue analysis (Fig. 3). In contrast, the normalized peak intensities of other elements (Na, H, Na, K, O, and Cl) did not clearly differ between melanoma and dermis for both pellet and tissue analysis.
3.2. Univariate analysis of Mg and Ca
For comparison of signal intensities between melanoma and dermis, the intensities of emission peaks of normalized spectra were calculated by integrating the area under the peak after subtracting the background. Specifically, the background signal near the emission peak was linearly fitted and subtracted from the raw data. The shot-to-shot signal intensity at a measurement spot of the pellet and tissue samples showed different trends. In the case of pellet sample, the signal intensity increased gradually with shot number, Fig. 4(a) , whereas that of the tissue sample showed little change with respect to shot number as shown in Fig. 4(b). The reason for the observed difference may be related to the sample preparation. The tissue sample was collected from the mouse and used for LIBS measurement with no pretreatment. On the other hand, the pellet samples were prepared by pressing dried tissues at the pressure of 1 ton. When the solid pellet sample was ablated by laser, the crater was expected to form a clear hole which then might have produced a confinement effect, the enhancement of signal intensity in contained plasma, resulting in an increase of signal intensity with increasing crater depth.
Figure 5 presented averaged intensities of Mg (II) and Ca (II). The results for pellet analysis, Figs. 5(a) and 5(b), showed that the normalized intensities of Mg and Ca were higher in melanomas than in dermis, with little variation. The low variation observed in pellet analysis may be explained by the homogeneity of pellet samples and similarity in water contents (i.e., freeze-dried). In excised tissue analysis, Figs. 5(c) and 5(d), the normalized intensities of Mg and Ca were also higher in melanomas than in dermis. However, there was more intensity variation within samples observed in the excised tissue analysis, which could be related to tissue inhomogeneity and differences in water contents, surface rigidity, and individual mice.
Based on these results, we concluded that the normalized intensities of Mg and Ca were higher in melanomas than in the dermis. These results could be explained by the increased numbers of cells and cell organelles induced by rapid proliferation of melanoma cells. For example, the endoplasmic reticulum and Golgi complex are both known to store Ca [12, 13 ]. Furthermore, melanoma cells have large amounts of melanin-containing pigment granules, which are rich in Ca . Therefore, high intensities of Ca in melanoma lesions may be explained by the increase in cellular components and organelles. Consistent with this, previous studies have reported high levels of Ca in cancer tissues, such as breast cancer, colorectal cancer , uterine cancer , and canine hemangiosarcoma . Mg has an important role in the biosynthesis of proteins that play a crucial role in cell proliferation . Furthermore, more than 300 different enzyme systems depend on Mg for their catalytic functions. High levels of Mg have also been reported in colorectal cancer  and breast cancer .
3.3 Multivariate analysis of Mg and Ca
The differentiation of melanoma and dermis can be done with improved accuracy when the differences in spectral intensities are considered simultaneously. For precise discrimination with a single shot LIBS spectra, statistical analysis was performed on the total spectra from all melanoma and dermis samples. Based on the univariate analysis results, the spectral regions surrounding the Mg and Ca lines, which showed quantitative differences between melanomas and the dermis, were used for the statistical analysis (pellet: Mg (II) 279.553 + 280.170 nm, Ca (II) 393.366 nm, Ca (II) 369.847 nm; tissue: Mg (II) 279.553 + 280.170 nm, Mg (II) 285.213 nm, Ca (II) 315.887 + 317.933 nm, Ca (II) 393.366 nm, Ca (II) 369.847 nm, Ca (II) 422.673 nm, Ca (II) 558.879 nm). First, principal component analysis (PCA) was used to reduce the high dimensionality in the data from all the pellet and excised tissue samples. Figure 6 showed a scatter plot of PCA scores on the derived variables (or principal components) of melanoma and dermis in pellet and excised tissue analysis. The first and second principal components in the pellet analysis accounted for 76.56% and 10.51% of the total variance, respectively. The PCA results of pellet showed that 92.63% of total variance could be represented by principal components 1–15. In contrast, in excised tissue analysis, the first and second principal components accounted for 73.48% and 12.85% of the total variance, respectively, and almost all of the variance (97.32%) was accounted for by principal components 1–15. Since the contribution of principal components other than 1–15 was very small, the additional components were ignored. The results of PCA, representing the data of melanomas and the dermis from principle components 1–15 were then used to perform classification analysis using linear discriminant analysis (LDA), which used 10-fold cross-validation. Computation of PCA and LDA was achieved using the Matlab statistical and bioinformatics tool box.
The results of multivariate analysis are summarized in Table 2 . The confusion matrix (tabular data for predicted class versus actual class showing the performance of a method) in Table 2 revealed that high classification performances were achieved in the analysis of both pellet and tissue samples. The sensitivity ( = TP/(TP + FN)), the proportion of true positives those were correctly identified by LIBS among the actual positive (or melanoma) samples, and the specificity ( = TN/(FP + TN)), the proportion of true negatives correctly identified among the actual negative (or dermis) samples, values from pellet analysis results were 0.994 and 1, respectively, whereas those from excised tissue analysis results were 0.967 and 0.997.
3.4 Clinical perspectives
Mohs surgery is widely used for complete resection of skin tumors using histological analysis of the margin. Histologically identified cancer-infiltrated areas of the resection margin are examined and removed again until no infiltration of cancer cells is identified. However, repeated excision and histologic analysis is a time-consuming process, both for the pathologist and the surgeon . Therefore, the LIBS system may be helpful to identify the surgical margins of cancers using a drilling method, thereby reducing the burden of repeated excision and histologic analysis.
Laser treatment has recently been used in recurrent cutaneous melanoma for palliative purposes [20, 21 ]. During such treatments, lesions are ablated without additional histological analysis. Thus, the LIBS system can assist in the removal of the lesion by providing information on the ablated lesion.
Several obstacles, such as problems with beam and detector position and space limitations, must still be overcome for practical use of LIBS in the clinical setting. However, skin is exposed to the external environment; thus, addressing these challenges and adjusting for these technical difficulties will be feasible. Indeed, application of LIBS in the clinical setting will be best achieved for procedures involving the skin.
An ablation and real-time detection system using LIBS for discrimination of melanoma from the surrounding dermis would be highly advantageous. In this study, LIBS analysis of pelletized samples and real skin of melanomas and the surrounding dermis collected from melanoma-implanted mice demonstrated the feasibility of Mg and Ca as biomarkers for discrimination. Statistical analysis using PCA and LDA also showed that this method had high sensitivity for detecting melanoma. These results supported the potential practical applications of this tool.
This research was supported by the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning (2014-049289, NRF-2014R1A2A1A11052300), and the Basic Research Projects in High-tech Industrial Technology Project through a grant provided by GIST in 2015.
References and links
1. S. A. Boppart, J. Herrmann, C. Pitris, D. L. Stamper, M. E. Brezinski, and J. G. Fujimoto, “High-resolution optical coherence tomography-guided laser ablation of surgical tissue,” J. Surg. Res. 82(2), 275–284 (1999). [CrossRef] [PubMed]
3. C.-S. J. Chen, H. Sierra, M. Cordova, and M. Rajadhyaksha, “Confocal microscopy-guided laser ablation for superficial and early nodular Basal cell carcinoma: A promising surgical alternative for superficial skin cancers,” JAMA Dermatol. 150(9), 994–998 (2014). [CrossRef] [PubMed]
5. S. J. Soong, R. A. Harrison, W. H. McCarthy, M. M. Urist, and C. M. Balch, “Factors affecting survival following local, regional, or distant recurrence from localized melanoma,” J. Surg. Oncol. 67(4), 228–233 (1998). [CrossRef] [PubMed]
8. A. N. Bashkatov, E. A. Genina, and V. V. Tuchin, “Optical properties of skin, subcutaneous, and muscle tissues: a review,” J. Innov. Opt. Health Sci. 4(1), 9–38 (2011). [CrossRef]
9. J. M. Waller and H. I. Maibach, “Age and skin structure and function, a quantitative approach (II): protein, glycosaminoglycan, water, and lipid content and structure,” Skin Res. Technol. 12(3), 145–154 (2006). [CrossRef] [PubMed]
10. E. F. Bernstein, C. B. Underhill, P. J. Hahn, D. B. Brown, and J. Uitto, “Chronic sun exposure alters both the content and distribution of dermal glycosaminoglycans,” Br. J. Dermatol. 135(2), 255–262 (1996). [CrossRef] [PubMed]
11. A. El-Hussein, A. K. Kassem, H. Ismail, and M. A. Harith, “Exploiting LIBS as a spectrochemical analytical technique in diagnosis of some types of human malignancies,” Talanta 82(2), 495–501 (2010). [CrossRef] [PubMed]
13. S. Chandra, E. P. Kable, G. H. Morrison, and W. W. Webb, “Calcium sequestration in the Golgi apparatus of cultured mammalian cells revealed by laser scanning confocal microscopy and ion microscopy,” J. Cell Sci. 100(Pt 4), 747–752 (1991). [PubMed]
14. W. D. Bush and J. D. Simon, “Quantification of Ca2+ binding to melanin supports the hypothesis that melanosomes serve a functional role in regulating calcium homeostasis,” Pigment Cell Res. 20(2), 134–139 (2007). [CrossRef] [PubMed]
17. M. Terasaki and H. Rubin, “Evidence that intracellular magnesium is present in cells at a regulatory concentration for protein synthesis,” Proc. Natl. Acad. Sci. U.S.A. 82(21), 7324–7326 (1985). [CrossRef] [PubMed]
18. P. M. Santoliquido, H. W. Southwick, and J. H. Olwin, “Trace metal levels in cancer of the breast,” Surg. Gynecol. Obstet. 142(1), 65–70 (1976). [PubMed]
19. A. Nijssen, T. C. Bakker Schut, F. Heule, P. J. Caspers, D. P. Hayes, M. H. Neumann, and G. J. Puppels, “Discriminating basal cell carcinoma from its surrounding tissue by Raman spectroscopy,” J. Invest. Dermatol. 119(1), 64–69 (2002). [CrossRef] [PubMed]
20. S. C. Gibson, D. S. Byrne, and A. J. McKay, “Ten-year experience of carbon dioxide laser ablation as treatment for cutaneous recurrence of malignant melanoma,” Br. J. Surg. 91(7), 893–895 (2004). [CrossRef] [PubMed]
21. J. A. van Jarwaarde, R. Wessels, O. E. Nieweg, M. W. Wouters, and J. A. van der Hage, “CO2 Laser Treatment for Regional Cutaneous Malignant Melanoma Metastases,” Dermatol. Surg. 41(1), 78–82 (2015). [CrossRef] [PubMed]