‘Err on the side of caution’: Screening for melanoma

Joshua Arbesman, MD, dermatologist and melanoma specialist at Cleveland Clinic, spoke with Healio about screening for melanoma. He highlighted those patients at the highest risk for melanoma development, the benefits and challenges of using telehealth for screening and the advantages of using advanced technology for screening and detection.

Healio: What patients are at higher risk for melanoma, and what are the screening guidelines for those patients?

Arbesman: There are definitely people who are at higher risk for melanoma development. Some of the factors we can see with our eyes and are evident, such as fair skin, a greater tendency to burn while in the sun or less ability to tan, light eyes, fair hair – red hair and blonde hair. All of these would be factors that increase the risk for melanoma, along with having an increased number of moles on your body – whether they are just normal moles and/or an increased number of atypical moles. Those are the general things we think about in terms of things we can see from a phenotypic perspective on an individual that would increase the risk for melanoma.

There are some genetic risk factors that may or may not be evident on the skin, that can increase your risk for melanoma. We are trying to explore more of those genetic risks factors so we can identify people who are at higher risk for melanoma development that we may not capture with those phenotypic characteristics that I just mentioned. The classic gene that increases your risk for melanoma development is having a mutation in a gene called CDKN2A. That often is associated with an increase in atypical moles as well, but there are some patients who do not have an increase number in moles but do carry a pathogenic variant in that gene. Those are the major risks factors we talk about, as well as having a family history of melanoma.

We screen people in higher risk categories at least once a year, typically. Depending on what occurs at that initial exam, I often tell patients, “let’s see how your skin looks today and then we make a decision about how frequently you will to come in.” What that means is if I see a lot of moles on an individual, where they may have difficulty monitoring their moles for changes at home over a longer time, instead of doing it once a year, let’s think about doing it twice a year if they are able. Some of this is more art than science and working with the individual and what their level of concern is regarding their melanoma development. That is how I think about it, starting with a once a year exam and then seeing if we need to increase the frequency.

Healio: What are the challenges of melanoma detection?

Arbesman: The wonderfulness, but also the challenge, of skin cancer detection is we can see a lot of things on people’s skin. We are able to see a lot of growths at an earlier stage than we may see them on an internal malignancy screening and that is great. The flipside to that is we are dealing during a lot of growths that sometimes can mimic melanoma, and sometimes some melanomas are more subtle in their presentation. So, I always think about comparing it to a colon cancer screening situation. During a colonoscopy, most growths are sampled compared with a melanoma screening, when we are looking at someone’s skin, I may look over 200, 300, 400 lesions on an individual and have to decide whether or not we are going to biopsy one. It makes it challenging.

We also do not want to miss something that may be more subtle in presentation. Some melanomas fit the criteria that we often give to patients: the ABCDE criteria. But some melanomas do not, and they are much more subtle and require the use of additional technologies to help us in that diagnostic dilemma.

That being said, we obviously err on the side of caution and will biopsy more. We do not want to be 100% on our melanoma diagnosis because that indicates we are missing melanomas. So, we err on the side of caution. We will biopsy some things that are concerning but do not end up being melanoma.

Along with that, there are a lot of patients with atypical moles. Sometimes they can mimic melanoma, clinically. That can be difficult, but we do have tools to help us in that type of diagnostic situation.

Healio: What are the advantages and disadvantages of screening for melanoma using dermoscopy, confocal microscopy or computer-aided image analysis?

Arbesman: Dermoscopy is a tool that has been around for a decent amount of time. We use polarized and non-polarized light to look at a magnified view of a lesion, which gives us additional features that can help us diagnose melanoma. The advantage of it is that it gives us more information, but there is a learning curve to be more skilled in terms of interpreting that greater amount of data.

Along the same line, we can do serial dermoscopy imaging over time. We may take a serial dermoscopy image of a mole at one visit and compare it to what mole looks like over time. What that gives us is greater detail to whether that mole is changing over time or not, which we generally pick up earlier under dermoscopy than with the naked eye. Often, we detect the change earlier and can biopsy the lesion at an earlier point in its melanoma evolution.

Along with that, we can use total body photography, which is a naked eye tool. But it gives us that serial change over time. The flip side is that when we don’t see a change over a certain time, we can be more reassured that this lesion is not likely to change and is more likely to be benign. Oftentimes, we are talking about increasing the number of melanomas we are detecting at an earlier stage, but it is also important to decrease the number of benign lesions that are biopsied so that we are judicious in what we biopsy.

Reflectance confocal microscopy is not something that I use in clinical practice, some individuals with pigmented lesion clinics do use it. It provides a certain type of image that, like dermoscopy, has a ramp up in terms of information you have to learn and also takes some bit of time to set it up and interpret the image. It does increase your ability to diagnose melanoma and other lesions, but it just is that balance of time and cost for that individual visit.

We are learning more about computer aided image analysis. The wonderful thing is, as computers become more sophisticated, they will probably help us diagnose melanoma. The challenge up until now is ensuring the computer is diagnosing the right lesion. We see a lot of papers coming out that are saying the computer analyzed whether or not a lesion was melanoma as well or even better or close to the level of a trained dermatologist. The key point being that it was put on the right lesion. But as we talked about, dermatologists are analyzing 200, 300, 400 lesions and we cannot have the computer, necessarily, analyze all of them until we have the computer analyze the whole body at the same point of time. That will hopefully be in the future, and people are looking at that from a research perspective, and I hope it is in concert with what we do with our patients, helping us diagnose melanoma at an earlier stage and not over-biopsying as much as possible.

Healio: What is the clinical importance of a 31-gene expression profile (31-GEP) test, and when would you use it in your practice?

Arbesman: This is a commercially available test that gives us more information about the prognosis for an individual melanoma. Castle BioSciences, the company that makes the 31-GEP test, divide the melanomas into Class 1 or Class 2 and then they subcategorize within those classes. Class 2 being a worse prognosis, meaning the patient is potentially going to do worse, while Class 1 is potentially that the patient will do better.

The advantage is that it gives us more information about how the melanoma is going to do, but they are still gathering more data to narrow that down. The big question, which we are starting to understand a little better, is what do we tell the patient in terms of how we are going to manage them differently when they have a Class 2 result on that test, do we screen them differently – in terms of imaging tests, or do we think about whether or not they would be eligible for adjuvant therapy to prevent recurrence of their melanoma. That is what we are still trying to understand a little better.

If we tell a patient they are going to do worse, I would rather tell them that and tell them we are going to do something to help improve their prognosis. But if we do not have anything to do in that situation, it does not necessarily help the patient. It just increases their anxiety. We are really understanding better how we can help the patient, especially with the advent of new adjuvant therapies that are clinically beneficial.

At the Cleveland Clinic, we work as part of an interdisciplinary team. I, as the dermatologist, work in concert with our surgeon and oncologist to manage patients that are at high risk for melanoma. I have not personally ordered the test, but our surgeon orders the test because it helps determine where we go next. Sometimes, it can affect the imaging that those patients receive as well. So, I work in concert with them, but I do not necessarily order the test myself. Oftentimes, patients are going to the surgeon after me. That helps screening for where they go next.

Healio: How have telehealth options impacted screening and detection?

Arbesman: We are seeing a lot more about telehealth, in terms of screening, with COVID-19. We are understanding better about how that works. Many patients send us images of lesions they are concerned about. There are two aspects that are important to consider as we move into understanding how to better use this technology. One, is the quality of the images patients are sending us, and also, we do not have access to those other technologies, such as dermoscopy. I recently had a patient who sent in an image of a larger brown lesion on his scalp that he was not sure whether or not it was new. When he sent in the image, it was not a terrible photo, but I still could not tell from that image whether it was something to be concerned about. I had just seen him a few months ago, but he had a history of recent melanoma diagnosis and he was concerned. We were not sure when we could bring him in, but I was recently able to see him in person and quickly, with the use of dermoscopy, I was able to tell it was benign and we did not have to be concerned. The added technology is great, but with the caveat that we sometimes need dermoscopy photos – which patients are unable to provide themselves. So that is the first point.

The second point goes along with what I said about computer-aided diagnosis. The patient is still selecting the lesion. They often will select something that is visually concerning to them, but what can happen outside the view of the photo is a lesion that may be concerning to me, but not as concerning to them. We do not necessarily have the photo, and that is always the challenge. ... It is something that needs to be kept in mind as we increase the telehealth offering for skin cancer detection.

Also, if we tell patients we are concerned, we want to make sure they come into the office. There is always the chance that we tell someone we are concerned about a lesion, but then they do not come in, vs. if they already came in for the screening we can talk with them about what we need to do next. It is much easier in that context to help them.