Diagnostic imaging: State-of-the-art techniques advance ability to detect melanoma

January 1, 2009

The future of melanoma diagnosis will involve increasingly sophisticated devices such as multispectral digital dermoscopy, confocal laser scanning microscopy and perhaps optical frequency domain imaging, an expert says.

Key Points

Las Vegas - Newer imaging techniques, including multi-spectral digital dermoscopy, confocal laser scanning microscopy (CLSM) and, perhaps one day, optical-frequency domain imaging, likely will improve melanoma diagnosis in the future, an expert says.

The MelaFind (Electro-Optical Sciences) takes dermoscopic images at 10 wavelengths, ranging from 430 nm to 950 nm, then uses a sophisticated proprietary algorithm to analyze several features in comparison to databases of melanomas and nonmelanomas, says Arthur J. Sober, M.D., professor of dermatology, Harvard Medical School, Boston.

In this way, it determines whether a particular lesion is suspicious for melanoma.

In ongoing phase 2 trials with an interim system, he says MelaFind demonstrated considerably higher specificity than did expert dermatologists who didn't use the technology.

Studies

More specifically, in one study involving 352 pigmented skin lesions, MelaFind achieved 48.4 percent specificity versus 28.4 percent for expert dermatologists (p<0.0001).

In another study, these figures were 45.1 percent and 20 percent, respectively (p<0.0001).

Dr. Sober says, "Another way of looking at that is, how many pigmented lesions would you biopsy to find one melanoma?"

In this regard, MelaFind posted an over-biopsy rate of 5.7:1 versus 7.9:1 for dermatologists in the first study and 5:1 versus 7.3:1 in the second, Dr. Sober says.

"This means you would not be doing as many biopsies to find the same number of melanomas. So, it should spare patients from excessive removal of benign lesions," he says.

Pivotal trial

At press time, the system was about to undergo a pivotal trial at seven U.S. clinical sites. The goal was to accrue patients with a total of 93 melanomas.

"The study's objective is to have the machine suspect at least 92 of the 93 melanomas and to have its specificity be better than that of the expert dermatologists' clinical impressions," Dr. Sober says.

Electro-Optical Sciences plans to complete the study's enrollment by around July 2009.

While the MelaFind remains under clinical study, Dr.Sober says, "The CLSM (VivaScope, Lucid) is commercially available. It's a noninvasive way of looking at the skin at magnification, and interpretation is based on the skill of the observer."

Prospective study

A recent prospective study compared the diagnostic accuracy of in vivo CLSM versus that of dermoscopy for benign and malignant lesions.

In this study, researchers evaluated 125 patients with suspicious pigmented lesions using dermoscopy and CLSM, and confirmed lesion diagnoses through biopsies.

Out of 37 melanomas in the sample population, dermoscopy identified 33; CLSM identified 36 (Langley RG et al. Dermatology. 2007;215(4):365-372). Of 88 benign lesions, dermoscopy correctly identified 74, while CLSM found 73.

The study's authors concluded that dermoscopy and CLSM complement one another clinically and that no melanomas escaped detection when researchers used both technologies on each patient.

Looking to the future

Dr. Sober says optical-frequency domain imaging could prove helpful in melanoma diagnosis.

While the technology isn't presently used for this application, he says, "It provides impressively accurate images of vascularity. It's an example of where things are going with instrumentation" for possible use in melanoma diagnosis.

"With increased rapidity, manufacturers are introducing new pieces of equipment using physical principles that are doing things we didn't dream we could do 10 years ago," Dr. Sober says.

And in the next decade, he adds, there likely will be a relevant, easy-to-use instrument available for dermatologists' offices that will improve the diagnosis of melanoma.

"And in all likelihood, the devices that will appear when these technologies come into more general use will be easier to use and less expensive than the currently available ones," Dr. Sober says.

Disclosure: Dr. Sober is a consultant for Electro-Optical Sciences.

For more information: http://www.lucid-tech.com/