Introduction

Non-melanoma skin cancers (NMSC) are the most common cancers diagnosed in the western world ¹ . The continuing rise in the incidence of NMSC will translate to greater need for surgical interventions. This paper has aimed to review the current literature regarding non-invasive methods of investigation of skin lesions. Non-invasive diagnostic tools have received increased attention for diagnosis, screening and management of NMSC. Several modalities are commercially available. Most of these devices are still limited to use in tertiary referral centres and research facilities.

The gold standard for diagnosis of basal cell carcinoma (BCC), squamous cell carcinoma (SCC), and solar keratosis (SK) is biopsy and histological examination. However, multiple lesions occur in areas of chronic sun damage, such that in the majority of cases lesions occur in the context of ‘field cancerisation’. Here, the diagnosis is often made clinically without confirmation by histology, since repeated and multiple biopsies may not be practical or feasible. Since skin biopsy alters the original skin morphology due to iatrogenic trauma, inflammation and scarring, non-invasive methods allowing in vivo study of the skin are of great advantage.

The development of topical treatment modalities including Imiquimod, diclofenac, salicylates, Hyaluronic acid, 5-Fluorouracil and ALA-photodynamic therapy, have changed the management of NMSC. Topical treatment of lesions often does not incorporate histological diagnosis. As topical treatments are only indicated for superficial and ‘low-risk’ lesions ₁ , with further development of these treatment modalities the need for non-invasive diagnostic methods will likely increase.

In the past ten to fifteen years, a number of non-invasive diagnostic tools have been developed and been investigated for their applicability for screening, diagnosis and management of skin cancer. These modalities allow the examination of large affected areas and offer the potential for non-invasive monitoring of topical treatment modalities in NMSC. Technologies vary considerably with regard to their penetration depth, resolution and clinical applicability, and a number of studies have evaluated their diagnostic accuracy and sensitivity and specificity rates. While dermoscopy is widely used in the clinical setting, the application of others often remains limited to specialized cancer centres and research facilities.

A review of all literature using databases of Pubmed and Medline, searching for keywords of ‘noninvasive’ or ‘non-invasive’ or ‘minimally-invasive’, and ‘skin’ was carried out. All the titles and abstracts of articles found were searched and relevant articles were selected. A further review of all the references mentioned in the selected studies was carried out and all relevant articles were added to the database. All selected articles were reviewed and categorised into groups based on the technique or the technology being investigated.

Dermoscopy

Dermoscopy uses horizontal pattern analysis for diagnosis by inspecting the skin with a hand-held lens using cross polarized light. The magnification reached by dermoscopy ranges from 6 to 100 fold, depending on the device. Penetration depth reaches the level of the papillary dermis ₂ . It is used widely in dermatological practice for the differentiation of melanocytic lesions, and sensitivity and specificity rates of 73%–96% and 73–100% have been reported for detection of melanoma ₃ . With respect to NMSC, studies have focused on the differentiation of pigmented BCC and SK against melanoma whereby dermoscopy may aid in the differential diagnosis ₄ . Case reports have described dermoscopic features of facial non-pigmented SK and the dermoscopic differentiation of superficial BCC and SCC in situ ₅ . Dermoscopy adds another dimension to clinical diagnosis due to better magnification and its use can increase accurate diagnosis of lesions of most types by surgeons and other clinicians.

High Frequency Ultrasound

HFUS uses ultrasound of frequencies between 20–100 MHZ to evaluate skin morphology. Images are obtained in vertical sections with penetration and resolution varying with the respective frequencies. 20 MHZ ultrasound has a penetration depth of 3.8 mm with an axial resolution of 39 µm and a lateral resolution of 210 µm. Newer devices employing 100 MHZ yield a resolution of 9.9 µm and 84 µm, yet have a decreased penetration of 1.1 mm ₆ . While routine ultrasonography is widely used clinically for evaluation of cutaneous, adipose, lymphatic, and deeper tissues, and has been used for preoperative assessment of skin tumours, HFUS has not yet been established for NMSC diagnosis outside investigational settings ₇ . Advantages of high frequency ultrasound include the deep penetration, whereby measurements of tumour thickness may be obtained on vertical images. However, the resolution does not reach the cellular level and histological subtypes of skin tumours may not be distinguished ₈ . Use of fine ultrasound probes through hollow bevel-tipped needles is currently being investigated experimentally.

Optical Coherence Tomography

OCT is an imaging technique based on interferiometry. The principle is comparable to ultrasound, but instead of longitudinal ultrasound waves, infrared-light is used, yielding an axial and lateral resolution of approximately 15 µm ₉ and a penetration depth of 500–1000 µm ₁₀ . The images obtained by OCT are 2 dimensional, cross-sectional and have a lateral dimension of 4 to 6 mm. It has previously been shown that by using OCT the layers of the skin as well as adnexal structures and blood vessels can reliably be visualized ₁₁ . However, no cellular or subcellular details may be seen. The basement membrane cannot be distinguished, such that early tumour invasion cannot reliably be determined ₁₂ . Preliminary studies have described the features of NMSC, including BCC and SK and suggest that this technique may aid in the evaluation of NMSC ₁₃ . In a recent study ₁₄ OCT features in NMSC were identifiable, but SK and BCC could not be differentiated. OCT diagnosis was shown to be less accurate than clinical diagnosis, but high accuracy in distinguishing lesions from normal skin was obtained, crucial for delineating tumour borders.

Fluorescence Confocal Scanning Microscopy

Confocal microscopy is based on the detection of exogenous or endogenous contrast within the tissue. Both fluorescence and reflectance mode confocal laser microscopy have been evaluated for their clinical and investigational application in dermatology ₁₅ . The principle of FCSM is the excitation and detection of fluorophores by scanning conjugated horizontal planes within the tissue using a laser light source. Exogenous fluorophores are injected using a hypodermic syringe and image contrast is achieved by differential-distribution and accumulation in the intercellular and intracellular compartment. The commercially available FCSM (OptiScan Ltd., Melbourne, Australia) employs an Argon-ion Laser (488 nm) for tissue illumination and excitation, while emitted fluorescence is used to visualize the morphological details ^{21 22} . FCSM images are oriented horizontally with a lateral resolution of 0.5 to 1 µm and an axial resolution of 3–5 µm, at 488 nm, the penetration depth reaches 200 to 250 µm (level of the papillary dermis). FCSM enables the visualization of cellular and sub-cellular structures in vivo, with a resolution which is comparable to routine histology sections. However, injection of fluorescence dye is needed, which removes the technique from the realm of ‘non-invasive’ cathegory. Another limitation relates to the innate instability associated with handheld devices ₁₆ .

Reflectance Confocal Microscopy

RCM is based on the reflectance, scattering and absorption of monochromatic light by endogenous chromophores such as melanin, haemoglobin and other cellular microstructures. Images have a resolution at the cellular level comparable to routine histology ₁₇ .

In the past decade, advances have been made in imaging human skin by RCM in vivo. The main investigational effort, however, lies in the evaluation of melanocytic lesions and high sensitivity and specificity rates have been described. Repeated imaging can be performed, and the motion-artefacts of life imaging are overcome by stabilizing the objective lens with an adhesive ring device on the skin surface.

Of note is the disadvantage of studies that have used RCM in diagnosis of NMSC. The most cited studies regarding the use of RCM in diagnosis of BCC have been conducted by a single group ₁₈ . These studies all appear to use a study ₁₉ conducted in 2002, which had used 8 lesions in 5 patients, as a sentinel study for creating the diagnostic criteria for BCC, and measuring sensitivity and specificity levels, as the basis of reference for all future studies of the group _{20 21} . This introduces possible questions of bias in the integrity of such studies, as well as issues regarding possible statistical power of a study conducted on 5 patients. Studies that have been performed outside of this group, which continue to use the original group's results as a point of reference, have never been as extensive, or as clear in demonstrating the benefits of RCM ₂₂ . Apart from the studies conducted by that group, there is a lack of studies on the human skin performed by means of RCM.

The main limitation of the RCM technique is the failure to visualize the depth invasion of skin tumours due to its relatively shallow penetration on horizontal sections and inability to evaluate lesions with significant hyperkeratosis. RCM therefore, does not permit an evaluation of the basement membrane on horizontal sections, thus early invasion of SK and progression to invasive SCC cannot be determined. RCM may allow therapeutic monitoring of patients with SK, where previously only clinical evaluation was employed to assess efficacy with the potential to detect sub-clinical or residual disease. Thus patients receiving non-invasive therapy, e.g. patients with SK or BCC treated with Imiquimod may be evaluated using RCM permitting a systematic morphologic description of defined skin sites as treatment progresses ₂₃ , although these studies have used the criteria of BCC diagnosis that were developed by the same group of researchers as mentioned above.

Summary

Clearly, non-invasive methods of diagnosis on NMSC would provide major clinical and economic benefits for surgeons and the community and represent a significant step forward. A number of non-invasive diagnostic methods are currently being evaluated for their clinical applicability for NMSC diagnosis. The obvious advantages of non-invasive diagnostic tools are the lack of tissue processing or staining, the possibility of examining tissue in its native state, thus permitting repeated imaging or monitoring of selected skin sites over time. While dermoscopy is routinely used for evaluation of BCC, its role for SK may be limited to pigmented SK due to its unspecific features. HFUS presently remains an investigational device, requiring further clinical studies to evaluate its applicability in skin tumour management. OCT, while only used in specialized skin cancer centres, appears to have the potential for clinical applicability and evaluation of NMSC. For HFUS and OCT comparable systematic studies are lacking, and future investigations will have to determine their role in clinical dermatology. While the use of non-invasive imaging devices may aid in the diagnosis and management of NMSC, there are several limitations. Both HFUS and OCT lack the appropriate resolution. Further improvement of the described technologies may overcome these problems in the future. Our group and others are currently studying the potential role of RCM in improving in vivo diagnosis and better assessment of surgical margins for NMSC.