Diagnosis: Melanoma
Bookmark and Share

What emerging tests and tech mean for dermatologists

Early diagnosis of melanoma is critical in the battle against this deadly disease — with a 98-percent five-year survival rate if it is diagnosed and treated at an early minimally invasive stage, according to data from the government’s Surveillance, Epidemiology, and End Results Program (SEER). New technologies promise to help improve dermatologists’ efficiency and accuracy for early detection. Amid this promise, though, it is important for clinicians to recognize the strengths and limitations of available tests.

For example, if a molecular test provides negative results for a suspect lesion, the clinician must understand the sensitivity and specificity of the assay and how to incorporate a result in conjunction with the clinician’s clinical impression, said Pedram Gerami, MD, associate professor of dermatology and pathology and director of research at the Skin Cancer Institute of Northwestern University’s Feinberg School of Medicine.

Clinicians also need to understand how test results affect management. “They really need to know how to interpret the tests, the meaning of the tests, and the limitations,” Dr. Gerami said. “We have to keep up, or else other fields that are more in tune to the molecular era [tests], will take over the care of a lot of the patients that we currently manage.”

Identifying suspect lesions

“The gold standard for melanoma diagnosis is a full-skin examination and biopsy of suspicious lesions,” said Debjani Sahni, MD, assistant professor of dermatology and director of the cutaneous oncology program at Boston University School of Medicine. But detecting melanomas accurately, particularly early melanomas, remains a challenge for dermatologists, she said. “A lot of benign lesions get biopsied unnecessarily, causing morbidity to patients, who can end up with a lot of scars and be put at a psychological disadvantage. Moreover, early melanomas may still be missed on examination.” [pagebreak]

Possible adjuncts to the biopsy include imaging booths, dermoscopy, multispectral imaging, confocal microscopy, and electrical impedance spectroscopy.

Dermatologists often use digital photographs to detect lesion changes between visits, which can be a time-consuming process. However, a total-body imaging booth developed by Clara Curiel, MD, and her co-founder of DermSpectraTM, Karleen Seybold, may make it easier to photograph patients’ lesions, said Sancy Leachman, MD, PhD, chair and professor of the department of dermatology and director of the melanoma program at Oregon Health and Science University’s Knight Cancer Institute. If this becomes available, she said, it would enable clinicians to use the automated system to compare total-body photos from one visit to the next.

High-resolution whole-body images are acquired from the system in less than 10 minutes. They are standardized and repeatable over time, with optimized lighting and reproducible poses. The system allows comparisons of the images on a viewer, and the company plans to eventually automate change detection. The system images 85 percent of the body surface (without imaging the scalp or soles of the feet). The company explained that that beta testing units continue to be tested in dermatology private practices in Arizona, and it is expected that commercial units will be available in 2015. Early patient survey results demonstrate a high level of confidence in whole-body imaging as compared with hand-written notes in documenting change. There are plans to extend applications to other specialties and potentially for use in clinical trials.

More dermatologists are using dermoscopy, allowing them to use magnification and a polarized light to examine lesion features that cannot be seen with the naked eye. However, the utility of this tool for guiding biopsy decisions depends on the user’s level of experience with dermoscopy, said Laura Ferris, MD, PhD, associate professor of dermatology at the University of Pittsburgh. [pagebreak]

Studies including primary care physicians as subjects that were reviewed in a 2012 paper published in Canadian Family Physician indicated that one day of instruction and reference information can improve dermoscopy skills (58:740-745). These studies reported sensitivity ranging from 53.1 percent to 84.6 percent and specificity from 71.8 percent to 89 percent.

Another diagnostic technique, multispectral imaging, captures imaged data by illuminating the skin in different spectral bands in the visible and near infra-red (IR), Dr. Sahni explained. “This enables extraction of specific information from an object that the human eye may have failed to capture,” she said. One such device, the automated, handheld MelaFind® system (MELA Sciences), Dr. Ferris explained, examines a lesion with 10 different wavelengths, from 430 to 950 nm, and runs it through a computer program that automatically analyzes 5,000 different features within the mole and compares it with its database, providing a probability score indicating how atypical it is. Based on the probability score the device provides, the clinician can decide whether to perform a biopsy, Dr. Ferris said.

In a prospective, multicenter, blinded study published in Archives of Dermatology in 2011 that included 1,831 pigmented skin lesions from 1,383 patients, with 1,632 eligible lesions, MelaFind showed a sensitivity of 98.4 percent and biopsy ratio of 10.8:1 (147:188-194). When borderline lesions were added, the sensitivity was 98.3 percent and the biopsy ratio was 7.6:1.

Confocal scanning laser microscopy (CSLM) creates in vivo images of skin lesions. “A laser beam is used on the lesion of concern, and the reflected light from that lesion is transferred into an electrical signal that produces a 3D image. The resolution of the images is comparable to standard pathology,” Dr. Sahni said. “So it’s like getting a virtual biopsy of the lesions that you’re concerned about.” [pagebreak]

The technology is based on the reflective properties of the tissue, Dr. Sahni said, and early melanoma and benign lesions will have different reflective properties. Dr. Sahni explained that users need to be trained and findings are dependent on the user’s interpretation. In a study published in the Journal of Investigative Dermatology, five independent observers demonstrated a sensitivity of 88.15 percent and specificity of 97.60 percent when using this technology after receiving 30 minutes of training (2005;124:493-498).

A retrospective study of CSLM was published in the British Journal of Dermatology in 2008, in which observers examined 3,709 unselected CSLM melanocytic tumor images from 20 melanomas and 50 benign nevi. The study reported a sensitivity of 97.5 percent and specificity of 99 percent (158:329-333).

“What they’re still working on is obtaining reproducible, high-resolution, in vivo imaging that can be visualized in real time with the same (or better) level of diagnostic accuracy as you can get with a biopsy,” Dr. Leachman said.

Investigators are examining electrical impedance spectroscopy as a potential melanoma diagnostic tool, in which a minute electrical impulse would travel from one probe to another through the skin. Malignant cells impede the current differently than benign cells. “You’re using a different physical property, that is, something other than light-based visualization, to get at the same problem of differentiating benign from malignant,” Dr. Leachman said.

A multicenter study published in Experimental Dermatology reported a sensitivity of 95 percent for malignant melanoma and specificity of 49 percent when 210 lesions (148 benign lesions; 62 malignant melanomas) were examined with this technology (2011;20:648-652). [pagebreak]

Although many of these methods are sensitive in detecting suspect lesions, biopsy is very specific, said Kavitha Reddy, MD, assistant professor of dermatology at Boston University School of Medicine. “Because melanoma is such a serious disease, we’re still giving the patient the test that is virtually 100 percent reliable at this point,” she said. “Reduced specificity is why these other technologies aren’t mainstream yet, but technology is certainly advancing and I’m quite hopeful that, within a number of years, we will hopefully be doing noninvasive diagnosis.”

Taking a deeper look

Molecular assays and mRNA expression profiling provide additional information to assist clinicians in diagnosis, determination of prognosis, and treatment of melanoma.

Derm Tech International’s Pigmented Lesion Assay may enable clinicians to examine cells harvested noninvasively. If it is approved by the FDA, physicians will be able to apply an adhesive patch and remove it to collect mRNA from a pigmented lesion. “That mRNA is then evaluated using a technique called mRNA expression profiling, which basically looks at the pattern of expression of the mRNA and then classifies that lesion as benign or malignant,” Dr. Gerami said.

This tool could have a significant impact on the field, Dr. Gerami said, potentially eliminating the need for extensive biopsies in patients with dysplastic nevus syndrome or many moles, unless results indicate the need to biopsy the lesion.

Molecular assays performed on biopsy specimens have taken on expanded roles the last several years. “The ways that these tests are most commonly used are in diagnosing borderline lesions, lesions that look like melanoma under the microscope but could also be a benign mimicker of melanoma,” said Daniel Miller, MD, assistant professor of dermatology and dermatopathology at Boston University. These may include certain dysplastic nevi, deep penetrating nevi, and various types of Spitz tumors. [pagebreak]

Molecular assays may also provide important prognostic and therapeutic information by detecting specific mutations in the melanomas, which may enable physicians to prescribe targeted therapies, Dr. Miller said. “These tests do not replace a microscopic exam, as they provide prognostic, not diagnostic, information. The gene mutations do not diagnose a melanoma, but we use them to generate a profile that helps us decide what drugs might be useful for the patient,” he said. These assays can also help predict which melanomas may be more aggressive clinically.

The DecisionDx-Melanoma test (Castle Biosciences), an mRNA expression profiling test, can examine specimens removed previously, Dr. Gerami said. Traditional staging systems used to predict the aggressiveness of a melanoma are based on histologic parameters, such as the depth of the melanoma, mitotic activity, presence of ulceration, and sentinel lymph node biopsies. “Data thus far suggests that this molecular test, which they can do on the primary tumor, provides information beyond conventional staging methods. It’s done on tissue that’s already been sampled,” Dr. Gerami said. “Based on this molecular test, we can classify melanoma into those that are less aggressive and those that are aggressive and allow for more personalized care for melanoma patients.”

Fluorescent in situ hybridization (FISH) probes are used to evaluate DNA for copy number changes. “We use this method to evaluate a group of lesions that are considered borderline by histopathology,” Dr. Gerami said. In these cases, he explained, FISH can be used to look for gains and losses of certain critical parts of DNA to see if the pattern is more consistent with a melanoma or nevus and then make a more accurate prediction as to the behavior of the lesion. myPath Melanoma (Myriad Genetics), based on mRNA expression profiling, also enables clinicians to assess mRNA expression patterns in borderline or ambiguous lesions, Dr. Gerami said. [pagebreak]

Implications for clinical practice

Despite advances, experts believe current tools will not replace the need for experienced dermatologists. “They’re adjunctive tests that can greatly enhance the capability and level of care, but they’re not replacements,” Dr. Gerami said. In some cases, he explained, there is a high rate of false-positive and false-negative results, or tests may be appropriate only in certain circumstances. “In the hands of an expert user, the potential is great and the potential for impact in the field is really tremendous,” he said.

Primary care physicians remain important partners in the process. “Ultimately I feel the dermatologist is the key in making the diagnosis, but the primary care physician is integral in referring patients who are likely to be at a higher risk,” Dr. Sahni said. Therefore, dermatologists need to guide primary care physicians regarding taking a thorough patient history and checking for risk factors for melanoma, questioning patients on any changing moles, and knowing when to refer patients appropriately to dermatologists, she said.

Assembling the pieces

“There are technologies at every level that that are beginning to contribute to our understanding of what a melanoma is,” Dr. Leachman said. “Ultimately, what we’re going to have to do is put that whole spectrum of technologies together and learn how to useall of these technologies in conjunction with one another so that you build a complete picture. Then you’re going to need to validate the new and improved combination of technologies with long-term outcomes in order to prove it is a new gold standard.’”

In the end, Dr. Leachman said, new technologies can be useful tools. “You have to be able to adapt,” she said. “I think the people who are going to struggle the most are the ones who fail to figure out how to adapt and incorporate these new technologies into their everyday lives and everyday practice.”

Editor’s note: Dr. Gerami is a consultant for and has received honoraria from DermTech International, Castle Biosciences, Abbott Molecular, and Myriad Genetics. Dr. Leachman has worked with Myriad Genetics. Dr. Ferris has been a consultant and investigator for Castle Biosciences, DermTech International, and MELA Sciences. Dr. Sahni, Dr. Miller, and Dr. Reddy do not have a financial interest related to their comments. 

 
Want to hear more of the conversation? Visit www.aad.org/education/aad-professional-education/dialogues-in-dermatology and access the May 2014 issue.

Related Resources

Arming the patient