ABC News and other media outlets recently covered the story of the World’s Smallest Skin Cancer – diagnosed by a team at Oregon Health & Science University and noted by Guinness World Records. How is technology changing melanoma detection? What’s the future?
For expert advice, I reached out to Mohs surgeon Anthony Rossi, MD, assistant attending dermatologist at Memorial Sloan Kettering Cancer Center in New York.
How is technology changing melanoma detection?
There has been an onslaught of advances for melanoma detection. Some can be applied to skin cancer in general, but melanoma is the holy grail of skin cancer detection. The idea of early detection is very interesting because you want to balance early detection with overdiagnosis and too many biopsies. We want to improve that biopsy to diagnosis ratio.
There is now non-invasive imaging called confocal microscopy. It looks like ultrasound but uses light waves to see cellular details more clearly than what you would see with ultrasound. You can see the cells in real time, and the clinician can make a judgment of whether the cells look atypical or normal. This sounds run of the mill, but the advances to get us here are significant. We can actually see below the skin and see the details of the cell. Confocal is now used in diagnosing a variety of malignancies and also benign skin conditions. It’s making its way from the bench side to clinical practice.
What are your thoughts on dermoscopy and Reflectance Confocal Microscopy for melanoma detection?
Dermoscopy is the gold standard, which is why we call it the dermatologist’s stethoscope. When you see the patterns of moles, we can see benign vs. malignant features. For any skin exam, I’m always using a dermatoscope. It helps reduce the number of biopsies needed when the lesion may look atypical with the naked eye.
We can also use digital photography to take photographs of moles and the body, and understand which moles are not changing and which moles have changed in a short amount of time. For the most part, these have become standards in practice.
Confocal is very exciting. There is an acquisition cost to buy the device, so right now it’s not in most clinical offices, but mainly in academic centers. A great amount of studies have shown the utility of confocal. I always think of making a diagnosis as a pyramid or a building. If something looks suspicious with the naked eye, then use a dermatoscope. If a diagnosis still isn’t clear, then proceed to confocal. Use a stepwise approach.
Which melanoma detection technologies do you think need to be further studied?
There are other technologies, such as electrical impedance, that are still being researched. Electrical impedance determines the electrical resistance of tissue, which is interesting technology and a biological property of some malignant cells compared with benign cells.
Tape stripping is a technology that is becoming more commonplace. The clinician uses tape to strip off the stratum corneum of the lesion. The genetic information from the skin can be analyzed to look for mutations that could determine whether the lesion is malignant or benign. It’s non-invasive, which could be important for cosmetically sensitive areas. Taking a biopsy is not easy in some areas of the body nor would you want to leave a mark in those areas.
As people use tape stripping more we will have more data on utility. Tape stripping is not always great on the palms or soles, so more work needs to be done. Plus different malignancies have different genetic signatures. The more we test, the more we learn.
What are the drawbacks of using new technologies to diagnose melanomas?
If you use it on every lesion, know that there will be false positives and false negatives just as there are with every test. Don’t use carte blanche – use your judgment to determine what’s atypical and whether to biopsy or use ancillary tests.
AI is exploding and this is very cool. I like to call it augmented intelligence. It’s not meant to be used on its own without the guidance of a clinician. The algorithms, while intelligent, are not foolproof so you’re going to get false negatives and positives, which could be misleading. Use the technology to help make the call. Algorithms for melanoma are getting smarter as the algorithms get more images and learn more about melanoma.
What would you say to clinicians who do not have access to these new technologies but still want to catch melanoma in its earliest stages?
For the most part, dermatologists do really well at catching melanoma and skin cancer on a clinical exam. They do an excellent job with skin checks when they are thorough. Use of digital images is extremely helpful in patients with a ton of moles. Those images are going to tell you at one point in time they had all these moles and 3 months later something has changed. We call this “short-term digital monitoring,” and it’s super helpful. Photos can tell you if a lesion has changed. When you look at the ABCDEs, those are all very helpful indicators of melanoma, but remember that the E stands for evolution. If a patient verbalizes that a lesion is changing, I take that very seriously.
What do you think is the future of melanoma detection?
The future is this multimodality approach. We are already employing a lot in the research setting at Memorial Sloan Kettering. Patients receive a total body 3D scan, if applicable. This takes a 3D avatar photo of the body. All the patient’s moles are mapped for research purposes, and we have computer algorithms that analyze the moles.
The future is almost here – or it is here and it’s being tweaked. The algorithms are getting more sophisticated as there are more inputs, and this makes our job more sophisticated. I can really hone in on lesions we have been tracking with 3D imaging. This also makes for a better experience for patients. The algorithm is not always right, but it clues me to areas where I can hone in. It takes a human brain to synthesize the information, but we are getting there.
What else should clinicians know about reflectance confocal microscopy and other new technologies for melanoma detection?
Tape stripping is very cool as it uses genomics – gene expression profiling – which is improving as well. If we know some lesions’ genetic profile, it can tell you if the lesion is a good actor or a bad actor. We have so much data now – it’s an exciting time in medicine. The job then becomes how fast we can create tools to help predict how something will play out so we know if we should intervene.
Did you enjoy this Patient Buzz expert commentary? You can find more here.
