Page 35 - Read Online
P. 35

Page 2 of 14                                Liu et al. J Cancer Metastasis Treat 2019;5:4  I  http://dx.doi.org/10.20517/2394-4722.2018.55

               with amide I and amino acids, and a decrease in carotenoids and lipids. The PCA-SVM statistical analysis based on
               the molecular fingerprints of the biomarkers yielded a 93.0% diagnostic sensitivity, 100% specificity, and 94.5%
               accuracy compared with histopathology reports.

               Conclusion: VRR can enhance molecular vibrational modes of various native biomarkers to allow for very fast
               display of Raman modes in seconds. It may be used as a label-free molecular pathology method for diagnosis of
               skin cancer and other diseases and be used for combined treatment with Mohs surgery for BCC.


               Keywords: Visible resonance Raman spectroscopy, human skin, basal cell carcinoma, principal component
               analysis, supports vector machine, molecular fingerprints, tryptophan, carotenoids





               INTRODUCTION
               Several studies have reported that skin cancer is in an increasing trend worldwide. Each year, new cases of
               skin cancer have been reported to have a greater incidence than breast, prostate, lung and colon cancers
               combined. Basal cell carcinoma (BCC) is the most common skin cancer, accounting for about 80% of the
               non-melanoma skin cancers (NMSC) which includes BCC and squamous cell carcinoma (SCC). Before the
               age of 65, 40%-50% Americans will have BCC or SCC skin disease occurring. There are about 4 million cases
                                                [1-4]
               of BCC diagnosed in the USA annually . Although BCC lesions rarely metastasize and have a high curable
               rate, they frequently expand quickly and may not only become a cosmetic problem but infiltrate surrounding
               tissue causing functional problems. Therefore, early detection and rapid and accurate diagnosis become very
               important for treatment.

               Currently, clinical routine diagnosis of skin cancer is performed using biopsy and histopathology. The
               gold standard method requires freezing or reagent preparation of the biopsy tissue prior to microscopic
               analysis. This requires skilled technicians and expert histopathologists to perform the diagnosis, as well as
               considerable time before the results are available. This leads to a diagnostic accuracy of about 80% depending
                                    [5-9]
               on expertise and training . A new diagnostic technique that is more rapid with higher accuracy would be a
               great clinical advantage. A search is ongoing for a fast and accurate way for diagnosis while treating cancers.

               An optical biopsy (OB) technique would offer many advantages including in situ evaluating lesions. There
               is a panoply of OB techniques, such as label-free native fluorescence, Raman spectroscopy (RS), optical
               coherent tomography, and other optical imaging techniques for ex vivo and in vivo cancer detection in
               human tissues and cells. These techniques have advanced significantly since the earliest reports in 1984,
               1987, 1991 and 1992 by Alfano’s group [10-13] . In particular, OB techniques for ex vivo detection of a panoply
               of different cancer types such as NMSC have progressed in recently years. Some fluorescence spectroscopic
               techniques for diagnosis of BCC have been reported, such as fluorescence confocal microscopy that has
               been used to detect BCC skin cancer margin in Mohs excisions without the need of frozen and fixed section
               processing [14-20] . RS technique has some unique features such as using intrinsic biomarkers and operating in situ
               and in a time period of many minutes or seconds. This has led to a rapid progress in seeking commercial
               instrumentation for researchers and clinical applications in BCC and other skin cancer diagnoses [21-36] . Most
               of the reports demonstrated spectral differences between normal skin tissue and BCC lesions using near-
               infrared (NIR) laser excitation only. Some reported the use of high power and very long signal collection
               time (e.g., 300 mW of power and minutes of time) on cells and tissues [24,25] . The NIR Raman method which
               uses excitation light sources at 671 nm, 785 nm, 830 nm or 1,060 nm has limitations for practical applications
               described as follows: (1) NIR Raman method only detects non-resonant biomolecule components; (2) NIR
               laser source requires long excitation time. Even a confocal micro Raman system has poor signal-to-noise
               (S/N) ratio since the spectral peaks are weak and not enhanced. Increasing excitation light intensity and
   30   31   32   33   34   35   36   37   38   39   40