Available Literature

Our search revealed 51 abstracts on photography, representing 8.5 percent of the total number of abstracts reviewed (see Appendix D, Table D1). With respect to the types of skin cancer, 55 percent (28/51) pertained to melanoma diagnosis and 2 percent (1/51) to basal cell carcinoma. The other abstracts were on a combination of skin cancers, or skin cancer type was not specified (see Appendix D, Table D2). Of the 30 abstracts of primary studies, the most commonly reported outcomes were test accuracy (n=7), and lesion characterization (n=5). No current trials on photography were found in ClinicalTrials.gov.

A majority of the abstracts (60 percent) addressed information on photography features, device variations, algorithms/image classifications/checklists, privacy issues, diagnostic accuracy and training. Twenty-six percent of the abstracts included data assessing at-risk populations and treatment settings. The remaining abstracts addressed information on longitudinal followup and diffusion, as well as general introduction and view/opinion articles. Abstracts for photography lacked data assessing effectiveness for different racial/ethnic groups or information on safety/adverse events.

Description of Technique

The use of photography to capture specific suspicious skin lesions or the entire body skin surface for monitoring purposes is commonly used in dermatology practices, but not typically in a primary care setting. Three studies identified by our search assessed the utilization of photography in U.S. dermatology settings.^19-21 Technical advances in and the affordability and adaptability of digital cameras have rendered the use of film-based devices obsolete; a myriad of digital cameras and models are commercially available. In this review, photography refers only to those images that were captured or stored using digital technologies.

Total body photography (TBP), also known as whole body photography, surveillance photography, or total body mapping, involves the acquisition of clinical head-to-toe images of the entire skin surface. In TBP, a series of 25 to 40 segmental baseline images are captured.²² Images may be stored electronically and used for side-by-side comparisons at future visits, or may be analyzed by algorithms to make computer-assisted diagnoses of skin cancer. Copies of pictures may be handed to patients for assistance during skin self examination. The success of this technique depends on ease and standardization of image acquisition, quality of photographs, and the availability of photographs for clinical use.^23-25 Standardized poses and accessories like pose frames aid in minimizing non-lesional differences during follow up examinations.²⁴

Variations of Technique

• Total body imaging systems. These generally consist of a digital camera for image acquisition and a computer storage and retrieval system. They include the MIRROR™ DermaGraphix, FotoFinder BodyStudio LITE, and MoleMap surveillance program.
• Photography with automated image analysis. Several devices like MoleMax 1 Plus, DermAssist™, Molemax 3 and Melanoscan^® have built-in software that allow real-time comparisons of total body baseline and followup pictures, automatic mole counting to detect new lesions, and diagnostic algorithms.^32,39,40 Other techniques utilize 3D differential forms of skin surfaces to “realize automatic recognition of melanoma”⁴¹ or computer-based algorithms to evaluate pigmented skin lesions.⁴²
• Total body photography combined with devices for lesion evaluation.^16,34,43-45 An example of such a device is a dermoscope/camera combination (see Dermoscopy section for further details).
• Teledermatology. In this technique, acquired digital images are transmitted virtually through the Internet via mobile devices, such as 3G phones and personal digital assistants, and e-mail or specific Web applications.⁴⁶ The images are transferred either from other practitioners (teleconsulting) or directly from the patient to the clinician (telediagnosis).⁴⁶ Description of the use of digital photo images in developing dermatological diagnosis and medical management is available.⁴⁷ This technique confers the advantages to both physicians and patients by eliminating the need for clinic appointments or reducing long waiting lists for the receipt of results while maintaining expert management, although data privacy and physician training issues have been raised.
• The use of ultraviolet light photography, (e.g., Canfield Visia System), is discussed under Photodynamic Diagnosis, as it relies on the photodynamic properties of melanin in the skin.

Clinical Context of Use

TBP is recommended for screening patients at high risk of skin cancer (specifically melanoma).^48,49 High-risk patients are defined as those with more than 10 dysplastic nevi, a previous history of melanoma, a family history of melanoma in a first degree relative (parent, sibling, or child). However, the age of onset and frequency at which photographic surveillance should be performed is unclear.

In terms of setting, our technical experts also suggested that these techniques may be useful in the primary care setting and in rural areas with no access to specialists. Although TBP is widely used by dermatologists and oncologists in the U.S., it is not routinely used by primary care practitioners. The literature describes a wide range of practices of TBP. While some clinical practices have dedicated professional medical photographers, others depend on existing dermatologists, oncologists, or general practice physicians to use these devices. Various training modules on the use of specific devices are available on the manufacturers' Web sites.

FDA Status

The cameras used during total body photography are not considered medical devices. They are therefore not regulated by the FDA. Similarly, the imaging systems used to store, analyze, and transmit images are not under the FDA purview.

Summary

Overall, photography included 7 abstracts from 6 unique RCTs—accounting for 64 percent of all RCTs included in this technical brief (Appendix D, Table D1). The RCTs evaluated outcomes including diagnostic accuracy, excision rates, patient satisfaction, cost savings and treatment adherence and followup (Appendix D, Table D1). Outcomes were measured at both the patient level (5 trials) and provider level (2 trials), with study participant numbers ranging from 88 to 5784. Almost all of the trials were conducted in a primary care setting (2 in the United States); only one was set in dermatology. Four of the RCTs were conducted outside of the U.S.: 3 in Australia, 1 in the United Kingdom. Abstracts from non-randomized studies consisted of mostly comparative and non-comparative cohorts (47 percent)—the remaining abstracts included review articles and data from diagnostic test studies. We found that the use of photography to capture suspicious skin lesions of the entire body for monitoring purposes is commonly used in dermatology practices, but not typically in a primary care setting. Photographic surveillance is recommended for patients at high risk of skin cancer, based on family history, history of dysplastic nevi, or history of prior malignant lesions.^48,49 However, the age of onset and frequency at which it should be performed is unclear. The affordability and adaptability of digital imaging permit the increased ease of electronic image storage and allow for side-by-side comparisons at future visits. The evolution of computerized imaging systems has also enhanced the ability to convey these lesions from patients to providers and across provider types. The available data are limited on the role of photography in changing clinical outcomes, including confirmation that baseline photographs in specialty clinics improve the detection of melanoma, resulting in detection of earlier stage lesions or recurrent lesions. While there are some studies, principally from Australia, addressing the impact of photography in primary care settings, no similar studies have been conducted in the U.S. Furthermore, data are limited on the role of photography for specific racial/ethnic groups. No current trials on photography were found in ClinicalTrials.gov.

Available Literature

The majority of the included abstracts addressed dermoscopy (69 percent) (see Appendix D, Table D1). Our search identified 433 abstracts on dermoscopy from the following types of studies: 3 trials, 39 comparative cohort studies; 96 noncomparative cohort studies/case series; 52 diagnostic test reviews/studies; 49 case reports; 78 narrative reviews; 5 systematic reviews; 74 technical reports; 37 guidelines, opinion pieces, or commentaries, and others. Of these, 324 abstracts provided information on the type of malignancy studied; 238 on melanoma, 22 on BCC, 5 on SCC, and 59 were combinations. Three non-randomized studies were identified in the ClinicalTrials.gov registry (see Appendix C, Table C2).

The main topics covered in these abstracts were: (1) dermoscopic features including lesion characterizations and histopathological correlations (94 abstracts); (2) general introduction and how-to articles (71 abstracts); (3) digital dermoscopy including automation and computer analysis (42 abstracts); (4) dermoscopy algorithms/image classification/checklist (39 abstracts); (5) other aspects of digital dermoscopy including teletransmission of digital images (23 abstracts); (6) general diagnostic accuracy (24 abstracts); and (7) follow up studies to monitor the change in pigmented lesions (15 abstracts); and (8) training (19 abstracts). No more than 6 percent of the total abstracts reported on the following: (1) other technical aspects of dermoscopy; (2) guidelines or proposals; (3) dermoscopy in nonwhites; (4) pregnancy; (5) and other miscellaneous variables. For the 15 abstracts that reported on longitudinal follow up (ranged from 3 months to 4 years) using dermoscopy, the outcome of interest was mainly the change in the number and the characteristics of pigmented lesions. No change in survival outcome was reported.

Description of Technique

Dermoscopy (also known as surface microscopy or epiluminescent microscopy or dermatoscopy) provides at least a 10-fold magnification of skin lesions by using either nonpolarized or polarized light.⁵⁰ There is generally good agreement for overall dermoscopic patterns between polarized and nonpolarized dermoscopy (kappa 0.88 to 1.00).⁵¹ Differences between the two are detailed below. Dermoscopy is used to differentiate between benign and malignant pigmented skin lesions, and aids in the overall assessment of pigmented lesion morphology. Types of dermoscopy devices are as follows:

• Nonpolarized light contact dermoscopy.^14,16,52-54 This device uses a nonpolarized light source (a halogen light source at a 45° angle), and requires the use of an oil or gel interface on the lesion to prevent surface reflection. It provides better illumination and resolution than polarized dermoscopy. The colors of lesions appear sharper in nonpolarized dermoscopy compared with polarized dermoscopy; the former is therefore useful in visualizing milia-like cysts and comdeo-like openings, peppering, lighter colors, and blue-light areas. Its cost is approximately $150.00.
• Polarized contact/noncontact dermoscopy.^14,16,52-54 Polarized dermoscopy devices do not need a liquid interface and are equipped with a cross-polarized lens that absorbs scattered light waves. Polarized contact dermoscopy can attain the images of vascular and other deeper structures, and is a useful tool in visualizing melanin, blue nevi, and shiny white streaks. Polarized noncontact dermoscopy is better used for imaging mucous membranes. Since direct skin contact is not required for visualization, the use of noncontact dermoscopy minimizes the risk of nosocomial infection. These devices (contact or noncontact) cost approximately $300.00 or more.
• Combined polarized and nonpolarized dermoscopy.⁵³ These devices incorporate the desirable characteristics of both types of dermoscopy. Clinicians can choose to use either polarized or nonpolarized lights. Its cost is approximately $1200.00.

Variations of Technique

• Dermoscopy without image capture features.^14,66 The Dermlite^® handheld dermoscopic device is comparatively inexpensive ($300-$1000). Test accuracy varies depending on a user's experience. This device does not identify “featureless” or very early melanomas.
• Dermoscopy with image capture features.^14,66 These devices are equipped with a digital camera that captures dermoscopic images, and can store the digital images of pigmented lesions and identify changes over time.
• Dermoscopy with image capture features and analytical capability.^14,66,67 These devices are equipped with both a digital camera and computer software. They can extract and save clinical and dermoscopic information. Purported advantages are that these devices can be used by nonexperts, and they provide objective and reproducible results. Some of the systems provide computerized diagnostic results.

Clinical Context of Use

Dermoscopy may have different intended purposes depending on the clinical setting. In a primary care setting, dermoscopy could be used primarily to help a clinician decide whether to refer a patient's suspicious skin lesion(s) for dermatology consultation. In a dermatology setting, dermoscopy could be primarily used to help improve the diagnosis of melanocytic and non-melanocytic nevi and help monitor patients with multiple nevi.

Clinical settings in the abstracts reviewed were almost all based in dermatology offices or pigmented lesion clinics. Of the 400 plus abstracts, only seven were based in primary care settings.

A 2009 survey reported that 48 percent of U.S. dermatologists (1555/3209) are dermoscopy users (n=1555), while 52 percent are nonusers (n=1654).⁶⁸ Among 1555 dermoscopy users, the types of dermoscopy used are: polarized light noncontact dermatoscope (54.7 percent), nonpolarized light immersion dermatoscopes (30.0 percent), and polarized light contact dermatoscopes (21.8 percent).⁶⁸ Dermoscopy was principally used in the assessment of patients with pigmented lesions (70.7 percent of patients); the remainder of patients had nonpigmented lesions (28.6 percent ) or papulosquamous conditions (8.8 percent).⁶⁸ Another 2009 survey reported that 88 percent (81/92) of dermatology residents were using dermoscopy and the authors concluded that the use of dermoscopy has increased significantly during the last decade.²¹ One cohort study suggests that a dermoscopic followup program, tailored to the individual risk profile of a patient (e.g., familial atypical mole and multiple melanoma (FAMMM) syndrome, atypical mole syndrome (AMS), previous melanoma), would be effective in detecting melanoma.⁶⁹

One study evaluated the following factors associated with the use of dermoscopy: sex and age of dermatologists, teaching setting, years graduated from residency, and patients' geographic residence. ⁶⁸Reimbursement issues may limit its widespread use. Marchionda 2010 indicated that the lack of reimbursement from an insurance company would result in unwillingness to use dermoscopy among U.S. practitioners.⁵³

One cohort study on non-whites in Brazil evaluated the effectiveness of dermoscopy in individuals with darker pigmentation.⁷⁰

Diagnostic Accuracy

A total of 86 primary studies and five systematic reviews evaluated general and digital dermoscopy; specific dermoscopic image features; particular classification schemes and/or algorithms; teledigital dermoscopy; and/or computer-aided analyses for diagnostic accuracy.

One systematic review compared the diagnostic odds ratios for melanoma across the different algorithms of dermoscopy.⁷⁵ (Table 2) Three systematic reviews investigated the diagnostic accuracy of dermoscopy compared with naked eye examination for melanoma.^76-78 Most of the primary studies did not address the issue of potential verification bias as it was likely that only those patients with clinically suspicious lesions received biopsies. One systematic review examined the diagnostic accuracy of conventional dermoscopy compared with computer-aided dermoscopy for the diagnosis of melanoma.⁷⁵

Table 2

Algorithms used in dermoscopy.

Training to increase accuracy. Seven studies analyzed pre-post training in the use of dermoscopy to increase the accuracy of detection of melanoma. Most training programs were relatively short in duration (1 day to 2 weeks (1 hour per day for 2 weeks in a Web-based course)) and consisted of didactic sessions and/or interactive sessions with experienced instructors.

FDA Status

The following devices have received Class I FDA approval status: EpiScope^® Skin Surface Microscope (Model 47300) [Welch Allyn, USA; decision year 1992], NevoScope (TransLite USA; decision year 1996), Dermascope (American Diagnostic Corp, USA; decision year 1999), and MoleMax (Derma Medical Systems; decision year 1999). The following is a Class II device: microDERM^® (Visiomed AG, USA; decision year 2004).

Summary

Of the 431 abstracts reviewed in this brief, only three were RCTs. Almost all of the primary studies on dermoscopy were non-randomized. The non-randomized studies tended to focus on features of dermoscopic image that would be of diagnostic interest; digital dermoscopy and the use of computer-based analyses; and evaluations of different algorithms and classification schemes. We did not identify any controlled studies examining the use of dermoscopy to increase the detection rate of early stage melanoma. The primary studies that reported patient outcomes largely focused on number of new lesions and how lesions had evolved. No study reported on how the addition of dermoscopy affected survival from melanoma.

One RCT did compare dermoscopic evaluation and naked-eye examination in 73 primary care physicians in Italy and Spain and inferred the effect of the addition of dermoscopy on the likelihood that a primary care physician would fail to refer a patient with suspicious skin lesions for a second expert opinion. A second RCT of 913 patients in Italy examined the downstream effect on the number of skin lesion excised for diagnostic verification with the addition of dermoscopy in a pigmented lesion clinic.

Available Literature

Our systematic literature search of MEDLINE^® identified 72 abstracts relevant to confocal microscopy from the following types of studies: 17 narrative reviews, 12 technical reports, 7 diagnostic tests, 6 comparative cohorts, 26 noncomparative cohorts, and 4 case reports (see Appendix D, Table D1). Reported clinical settings included 14 dermatology, 2 primary care, and 1 oncology practice. Identified studies addressed the use of confocal microscopy in patients with suspected melanoma (n=29, 40.2%) and NMSC (n=15, 20.8%). Several studies (n=28, 39%) addressed its use in a combination of skin cancer types. (See Appendix D, Table D2) The most commonly reported outcome was lesion characterization (27 studies), followed by test accuracy (17 studies) (see Appendix D, Table D4). No clinical outcomes were identified.

We identified eight observational studies of confocal microscopy on the ClinicalTrials.gov registry (see Appendix C, Table C2). Five of these studies specified the use of reflectance confocal microscopy; the rest did not specify the type of technique. Although three studies were completed and one was suspended, results for these studies were not posted.

Topics covered in these abstracts included: (1) features of microscopic images histopathological correlates (36 abstracts); (2) general overview of the technology and its use (20 abstracts); (3) test accuracy including sensitivity and specificity data (10 abstracts); (4) technical report and glossary (3 abstracts); (5) diagnostic algorithms and automation (2 abstracts); and (6) other 2 abstracts). Out of the 36 studies that reported features of images and histopathological correlates, only 6 studies had more than 100 participants. All 10 studies that provided test accuracy data were done out of the US (6 in Austria, 1 in Australia, 1 in Germany, 1 in England, 1 in Sweden).

Description of Technique

Resolution of CSLM images is specific to each device, and is determined by the wavelength of the laser beam, the topical aperture of the lens, and the size of the pinhole.⁸⁰ The maximum depth of imaging is 350 μm. The uniqueness of CSLM lies in its imaging of not only the epidermis, but also underlying structures and the papillary dermis. With its high resolution, CSLM images can be evaluated in detail for the diagnosis of skin cancer and characterization of lesions.⁷⁹

Confocal microscopy images are in grayscale, therefore structures with higher reflectance are bright over a dark background. Standardized terminology for the evaluation of reflectance of confocal microscopy images was developed at an online consensus meeting between 2004 and 2005 and subsequently published.⁸² Although initial models of CSLM were bulky, hand held confocal devices are now available.

Variations of Technique

• Confocal scanning laser microscopy. This type can be in either reflectance or fluorescence mode. In reflectance CSLM, laser-illuminated tissue structures and melanin reflect light toward the confocal microscope detector.⁸⁸ It is more commonly used in a clinical setting, and can be either diffuse or polarized.³⁹ In fluorescence CSLM, a laser beam excites the endogenous or exogenous fluorescent molecules, which emit the signals to the confocal microscope detector.³⁹ Fluorescence CSLM is used primarily in research.
• One manufacturer, Lucid, Inc., produced three models of confocal microscopes. The newest model, called VivaScope^® 3000, is a handheld device, which overcomes the size limitations in the previous models. Lucid also developed the VivaNet^® Digital Imaging and Communications in Medicine, which allows storage and transfer of confocal microscopy images among healthcare providers in different geographic locations. According to a general review, Optiscan Pty. Ltd. also manufactures confocal microscopes, named Optiscan.™³⁹

FDA Status

We identified the confocal microscopy devices from both Lucid, Inc. and Optiscan Pty. Ltd. from the FDA CDRH database. Although the Optiscan Pty. Ltd. device has achieved FDA clearance, the intended use stated in its FDA approval summary was for use during endoscopic medical procedures. The series of VivaScope devices (Lucid, Inc, USA) and Optiscan™ have received the FDA Class II status; the former in 2008, the latter in 2010.

Summary

In this brief, no systematic review or controlled trial on confocal microscopy was found. Although observational studies describing the use of confocal microscopy exist, data from comparative studies with longitudinal followup among large populations are lacking. Additionally, test accuracy of this technology is yet to be formally demonstrated in the United States, despite some test accuracy data from Europe and Australia.

Available Literature

The literature we reviewed identified 34 abstracts relevant to Ultrasound/Laser Doppler, and Ultrasound in combination with other techniques found in a combination of radiological and clinical journals. There were 16 primary studies, comprised of 2 comparative cohort studies, 7 non-comparative cohort studies, and 7 diagnostic test reports. In addition, there were 18 reviews including 6 narrative review and 12 technical reports. (See Appendix D, Table D1). The most commonly reported outcome was lesion characterization (8 studies) with test accuracy being the second most common (5 studies) (see Appendix D, Table D4).

A review of the ClinicalTrials.gov database revealed only one study on the use of ultrasound. This study combined the use of ultrasound with laser Doppler, with the stated aim of early detection of metastatic melanoma (NCT00776945, accessed November 5, 2010). This observational study is scheduled to be completed in December 2014.

Description of Technique

Ultrasound with high frequency scanners of 20 to 50 MHz is useful as an adjunct in the accurate diagnosis of skin lesions.^39,89 High frequency ultrasound provides information on lesion quality and inner structure of tumors, based on different echogenic properties.⁸⁹ Current ultrasound imaging techniques allow for the three-dimensional C-mode (computed) scanning of structures in the skin in vivo. A review of the identified literature reported that ultrasound is primarily used in pre-planning for therapy and surgery (for examples see Guitera 2008,⁹⁰ Vilana 2009,⁹¹ Pellacani 2003⁹²) through its 3D imaging of malignant processes.^93-95 It is also used as an adjunct in the accurate diagnosis of skin lesions.^39,89 Two studies combining high-frequency ultrasound with dermoscopy^96,97 reported possible improvement in diagnostic accuracy (over sonography alone), and helpful information about tumor depth and location to assist in surgical planning. Ultrasound biomicroscopy (UBM), a technique generally used in the diagnosis of various eye abnormalities, has demonstrated preliminary usefulness in differentiating the histological components of cutaneous BCC and SCC,^98,99 and eyelid lesions.¹⁰⁰ For patients with BCC, High Frequency Ultrasound has also been explored to evaluate tumor margins.¹⁰¹

Variations of Technique

• Reflex Transmission Imaging (RTI). This is a particular form of high resolution ultrasound that can be joined with white light digital photography for classification of pigmented lesions. The RTI device, termed DermaScanC, reveals the vascularization of tumors seen with color Doppler sonography (B-mode). This technique may reduce the number of referrals for benign tumors without missing melanoma; however, the small number of studies assessing its use and expense may limit its utility.
• Color-coded duplex sonography. This technique involves coupling a B-mode (brightness) image with a pulsed wave Doppler, and provides data on blood flow in real time. One study identified the usefulness of this technique in its ability to distinguish between melanoma and other pigmented skin lesions;¹⁰⁶ two other studies demonstrated its potential as a prognostic tool for the identification of melanoma with high metastatic potential.^107,108
• Laser Doppler perfusion imaging. This technique is able to discriminate differences in perfusion levels between malignant melanoma and benign pigmented skin lesions. Vascularization of melanoma lesions has been a primary interest for researchers because of the hypothesized theory that vascularization gradually increases during the transition of a lesion from benign to dysplastic to primary melanoma.³⁹ Early studies regarding the use of this technique to differentiate between benign and malignant melanocytic skin tumors reported its usefulness as a discriminative adjunct in assessment; there were no abstracts found on this topic since 2004, nor were there any studies listed in ClinicalTrials.gov database.

FDA Status

No information was found on the FDA clearance status for the devices of this type on the FDA CDRH database for use in the evaluation of skin lesions.

Summary

In this brief, no systematic review or completed controlled trial on ultrasound or color Doppler technology was found. The available literature addresses the potential benefit of noninvasive ultrasound imaging of skin lesions as a source of important clinical information to improve accuracy of diagnosis and assist in pre-operative planning. The evidence accessed for this study indicated that, while it was first thought that ultrasound alone would be helpful in differentiating between benign and malignant lesions, research to date supports its use as an adjunct to other diagnostic tools, but does not provide support for its use as a stand-alone tool. Additional trials are needed in order to determine the value of ultrasound/color Doppler techniques in establishing the diagnosis of melanoma or NMSC. Information about training requirements, or evidence of effectiveness among different patient groups (history, race/ethnicity) was not identified.

Available Literature

Our systematic literature search on MEDLINE^® identified a total of 22 abstracts, 16 of which dealt with PDD for BCC, two for use with suspected melanoma, and four that addressed a combination of NMSC (n=2) or skin cancer type not specified (n=2) (see Appendix D, Table D2) These abstracts were reported principally in technical journals, rather than clinical journals. These abstracts included six primary studies, including one RCT, four comparative cohort studies and one single case report. The remainder of the studies were narrative reviews (n=9) or technical reports (n=7) (see Appendix D, Table D1). The single RCT, reported from Sweden, was designed to evaluate the tolerance threshold of four different application times of 5-aminolevulinic acid (ALA) in 40 patients (10/group). The endpoint of the study was the fluorescence intensity between normal skin and tumor tissue. In the remainder of the primary studies, two reported on lesion characterization and three reported on test accuracy. (see Appendix D, Table D4). No clinical outcomes were reported.

We identified a single study of photodynamic diagnosis in the ClinicalTrials.gov registry (see Appendix C, Table C2). This study, not yet open for participant recruitment, is designed to evaluate the effect of the topical application of ALA on protoporphyrin formation among patients with NMSC. As noted in the section on confocal microscopy, confocal laser scanning microscopy will be employed as part of the study outcome assessment.

Description of Technique

Topical application of ALA has been shown to produce increased concentration of endogenous protoporphyrin IX (PpIX), which has high fluorescent yield.¹⁰⁹ Studies have shown that PpIX accumulates in skin tumors at a much higher concentration than in normal skin.^103,110

Variations of Technique

• Ultraviolet light photography. Ultraviolet light is absorbed by melanin. The theory behind this experimental technique is that illumination by ultraviolet light could reveal irregular pigment distribution, and therefore could be useful in defining the borders of melanoma (e.g., lentigo maligna melanoma).⁶⁶ It is unclear how widespread the use of this technique is in the dermatology community, as we did not identify any other abstracts related to this technique as used in the detection of melanoma.
• Polarized light photography. This method relies on the fact that reflected light has two components—one regular reflectance to reflect the skin surface morphology, the other “back-scattered” from within the tissue.¹²¹ It is useful in the assessment of skin surface morphology when the proper polarizing filters and techniques are used. It can be used in the assessment of dermal melanosis. It is not widely used for assessing skin pigmentation; Taylor 2006 and others have highlighted the limitations of polarized light photography in darker skinned persons with Fitzpatrick skin types IV, V, and VI.^22,121
• Other topical therapies, principally used in the treatment of nonmelanomatous skin cancer, such as imiquimod and 5-FU could be used in combination with photography to highlight skin cancers.

FDA Status

No information was found on the FDA clearance status for the devices of this type on the FDA CDRH database for use in the diagnostic evaluation of skin lesions.

Summary

In this brief, a single RCT of PDD was found in which the technical aspects of method were explored. Much of the extant literature addresses the technical aspects of the photosensitizers or the available different light sources. Although the available literature addresses the potential benefit of this method in directing or limiting potentially disfiguring biopsies for patients with nonmelanomatous skin lesions, there is little to no evidence to support the use of this method in melanoma given current data on test accuracy. Information about training requirements or optimum clinical setting was also not identified.

Multiphoton Laser Scanning Microscopy

Multiphoton laser scanning microscopy, also known as multiphoton fluorescence microscopy or multiphoton excitation microscopy, uses more than one photon excitation to illuminate endogenous fluorophores in skin tissues, which emits a fluorescence signal to be captured by a detector.¹²² Similar to CSLM, it uses laser beam and allows imaging of tissues beyond the superficial epidermis. Unlike CSLM, this technique does not use a confocal pinhole filter.^123,124 Evidence of the current application of this modality is sparse. Our systematic literature search identified three narrative reviews and two diagnostic studies of multiphoton microscopy or tomography (see Appendix D, Table D1).

We identified two registered cross-sectional studies that assess the use of this technology for skin lesion evaluation. Both studies are based in Taiwan and are recruiting participants (see Appendix C, Table C2). The only commercially available device for multiphoton tomography is DermaInspect^®, manufactured by JenLab in Germany (jenlab.de/DermaInspect-R.29.0.html). We could not determine the FDA clearance status for this device on the FDA CDRH database (see Appendix C, Table C1).

Electrical Bio-Impedance

Different biological tissues have different electrical impedance spectra. The spectrometer measures impedance in different frequencies (1 to 1000 kHz) as different frequencies reflect different tissue properties. Skin electrical impedance has been found to be statistically different depending on tissue types (e.g., impedance of benign pigmented nevi has been shown to be different from basal cell carcinoma).¹²⁵ One group of authors reported using the SciBase I noninvasive electrical impedance spectrometer (SciBase AB, Huddinge, Sweden) to measure impedance of different skin lesions. The use of electrical bio-impedance in the detection of skin cancer remains investigational at this time. The five abstracts on bio-impedance that we identified were all published before 2006 (see Appendix D, Table D1).

A proposed advantage of bioelectrical impedance is that the data generated from this technology can complement information from visual inspection, and help prevent misdiagnosis of basal cell carcinoma and other types of skin cancer.¹²⁶ Even though statistically significant differences in impedance were found between tissue types in Aberg 2003,¹²⁵ the degree of overlap and within group variance were too high to allow for easy clinical differentiation based on impedance measurements.

A search of the ClinicalTrials.gov Web site (accessed November 3, 2010) identified an international, prospective, non-randomized study that collected data for optimization of an algorithm to classify skin lesions using electrical impedance. This study has been completed, but the results have yet to be posted (see Appendix C, Table C2). A non-randomized study is currently recruiting participants to collect data on sensitivity and specificity of SciBase III electrical impedance spectrometer to detect melanoma and the data will be used to support a Pre-market Application to obtain FDA approval (NCT01077050) [see Appendix C, Table C1].

Optical Coherence Tomography

Optical Coherence Tomography (OCT) is an imaging technique—akin to an optical ultrasound—that utilizes reflected light to produce cross-sectional subcutaneous images of tissue at a resolution equivalent to a low-power microscope. This technique provides tissue morphology imagery at a higher resolution (smaller than 10 μm) than modalities such as MRI or ultrasound. OCT allows for instant, real-time sub-surface images of tissue morphology at near-microscopic resolution and requires no preparation of the sample/subject and no ionizing radiation.

Our search identified five abstracts^127-131 examining OCT's application to the diagnosis of skin cancer (see Appendix D, Table D1). Two abstracts summarized technical reports.^130,131 A1997 technical report describes OCT as a promising new noninvasive diagnostic imaging method for the visualization of morphologic changes of superficial layers of human skin.¹³¹ A 2005 technical report describes possible histopathologic correlates of dermoscopic structures identified using OCT.¹³⁰ Olmedo 2006¹²⁹ presents findings from a noncomparative cohort study of 23 patients (49 lesions) utilizing OCT to characterize basal cell carcinoma in vivo. The Mogensen 2009¹²⁸ narrative review described OCT as an “emerging imaging technology” that is “still evolving and continued technological development will necessitate an ongoing evaluation of its diagnostic accuracy.” Additionally, “OCT is being integrated in multimodal imaging devices that would potentially be able to provide a quantum leap to the imaging of skin in vivo”. Forsea 2010¹²⁷ investigates the “utility of OCT for the diagnosis of non-melanocytic, nonpigmented cutaneous tumors”. The comparative cohort study assessed 15 patients with clinical suspicion of epithelial cancers and precancers along with 7 control patients with inflammatory skin diseases. All patients had perilesional skin documented by clinical digital photography, contact dermoscopy with digital image capture and OCT—final diagnoses were certified by histology. Results demonstrated that OCT “appears as a promising method of in vivo diagnosis of early neoplastic cutaneous lesions”. Moreover, combining OCT and dermoscopy for lesion evaluation resulted in improved diagnostic performance when compared to clinical diagnosis, OCT or dermoscopy alone.

A recent search on the ClinicalTrials.gov Web site (accessed 11-3-2010) identified one observational study investigating the diagnostic value and possibilities of OCT in non-melanoma skin cancer. The study is currently recruiting participants (see Appendix C, Table C2). No information was found on the FDA clearance status for the devices of this type on the FDA CDRH database.

Tape Stripping

Tape Stripping is a noninvasive ‘biopsy’ technology used to analyze superficial cells harvested from pigmented skin lesions (PSLs) suspected of being early melanomas. Cells from the upper epidermis are stripped off using an adhesive tape, and RNA from the PSL is harvested and analyzed via ribonuclease protection assay (RPA) to differentiate malignancies on the basis of gene expression profiles. A 1992 study¹³² of 150 PSLs concluded, based on estimates of sensitivity and specificity of tape stripping for the diagnosis of malignant melanoma, that this method may be a helpful diagnostic tool when used in conjunction with ABCDE guidelines. DermTech International (www.dermtech.com) developed and patented the commercialized form of this technology. DermTech's Epidermal Genetic Information Retrieval (EGIR™) utilizes a custom adhesive film to collect surface skin samples. EGIR™ is reported to be quick and painless and can be applied to virtually any skin surface. To help increase diagnostic accuracy, EGIR™ allows for re-testing of lesions. In a 2011 study by Wachsman, reporting on the testing with an independent dataset, this classifier discerned in situ and invasive melanomas from naevi with 100-percent sensitivity and 88-percent specificity, with an area under the curve for the receiver operating characteristic of 0·955.¹³³

A recent search on the ClinicalTrials.gov Web site (accessed November 3, 2010) identified one non-randomized study, sponsored by DermTech International, assessing tape stripping for diagnosis of early stage melanoma. This study has been terminated (withdrawn per sponsor and investigator), and results have not been posted (see Appendix C, Table C2). No information was found on the FDA clearance status for the devices of this type on the FDA CDRH database.

Thermography

Dermatologic use of thermography involves measuring and mapping surface skin temperature through direct contact (via application of liquid crystal plates to a part of the body) or at a distance (utilizing a highly-sensitive medical infrared camera and sophisticated computer interface). A single narrative review¹³⁴ published in 1995 from the San Gallicano Dermatological Institute for Research and Care in Italy assessed thermography and its potential application in clinical and experimental dermatology. Among the topics reviewed was the clinical use of thermography as a diagnostic tool for cutaneous melanoma. The review reports that due to high percentages of false-negative results from studies in the 1980s,^135,136 the use of thermography as a stand-alone diagnostic tool for melanoma has diminished. However, thermography used in conjunction with thermostimulation (application of thermal stress on the skin to be examined) has allowed for better differentiation of melanoma from other types of pigmented lesions. A recent search of the ClinicalTrials.gov Web site (accessed 11-3-2010) identified one trial studying the application of infrared thermography to find skin lesions in patients with Kaposi's sarcoma—a topic outside the scope of this technical brief. No information was found on the FDA clearance status for the devices of this type on the FDA CDRH database.

Multispectral Imaging and Fully Automated Computer-Based Analysis

A fully automated device that has been reported in the literature is a device that captures multispectral images of a pigmented lesion in 10 bands, from blue to near infrared (MelaFind , MELA Sciences Inc, Irvington, New York). It uses automated image analysis and statistical pattern recognition to help identify lesions that should be considered for biopsy. This multispectral imaging system shows quantitative and more objective results compared with conventional dermoscopic analysis, which is qualitative and potentially subjective.³⁷ Diagnostic performance in a prospective, multicenter study of patients with at least one pigmented lesion scheduled for biopsy was recently reported.¹³⁷ This device is currently undergoing an FDA Premarket Approval review for use by dermatologists.