Melanoma testing: Progress in practice
by Brook Brouha, MD, PhD
Until recently, a physician’s ability to interpret and predict the future of melanocytic lesions was limited to clinical observation and microscopic observation. Demand for an accurate, noninvasive diagnostic modality inspired a paradigm shift that led to the development of various prebiopsy tools.
Dermoscopy, confocal microscopy and computer-aided image analysis emerged into practice; however, research shows these tools can be subject to limitations of pattern and image recognition.
More recently, gene expression testing is being used for the identification of high-risk melanocytic lesions. Noninvasive gene expression testing (2-GEP) is a way of identifying high-risk pigmented lesions before unequivocal clinical characteristics, the current gold standard, have developed.
In my practice, I have found noninvasive gene expression testing is useful in the many cases in which the lesion falls somewhere in between obviously benign and highly suspicious. These are the patients for whom objective genomic testing is particularly helpful. Surgical biopsies remain the standard of care for melanoma diagnosis; however, there are numerous instances when we want to avoid unnecessary cutting.
Noninvasive genomic rule-out test
Noninvasive gene expression testing with the Pigmented Lesion Assay (PLA, DermTech) uses an adhesive patch for sample collection and provides information on molecular risk factors that cannot be visually ascertained. The assay evaluates how RNA levels of key genes in a particular lesion are changing and looks at tumor biology rather than what tumor cells look like.
A skin sample is collected with tape-like adhesive patches and sent to a CLIA-certified and CAP-accredited high-complexity laboratory. RNA levels from the sample are then analyzed for the detection of two key target genes — long intergenic non-protein coding RNA 518 (LINC00518) and preferentially expressed antigen in melanoma (PRAME) — overexpressed in melanoma. If one or both of the target genes are detected, the test is positive. About 10% of tested lesions are positive for the evaluated molecular risk factors, and clinicians biopsy 2-GEP positive lesions to determine next steps. Notably, 90% of lesions are negative by 2-GEP, and surgical biopsies and follow-up excisions are avoided. The 2-GEP provides an additional option to obtain objective information to guide clinical treatment decisions, and the negative predictive value of more than 99% adds confidence.
As a dermatologist, I have been looking at moles for many years. Hence, I have a confident feeling about the look of a clearly benign pattern. At the other end of the continuum, certain lesions are clearly melanoma. Unfortunately, not all pigmented lesions fall into these two simple categories.
I use 2-GEP principally for lesions that give me reason for concern because they have a suspicious characteristic or for lesions in which the patients express unusual concern. Patients have a unique lens over time that my office visit snapshot lacks. I take it seriously when a patient is worried about a “weird” or “changing” mole. On the other hand, if I am 80% sure I am looking at a melanoma, I go directly to surgical biopsy.
This is a helpful tool whether you are an experienced dermatologist or just starting out because it serves as a safety net for clinical decisions in the high-stakes world of pigmented lesions. For a physician, missing a melanoma can be a life-changing event, and even more so for the patient.
Case 1: Diagnosis of melanoma in situ
A 74-year-old woman presented with hair loss. She had a family history of melanoma (brother) but no personal history of melanoma; a personal history of nonmelanoma skin cancer was reported. After hearing about her multiple sunburns and seeing the sun damage on her face, I suggested she undergo a full skin exam. The exam revealed significant sun damage and several well-healed scars from past basal cell carcinoma (BCC) and squamous cell carcinoma (SCC) excisions. I noticed a 6-mm variegated black and brown macule on the patient’s left shoulder (Figure 1).
She could provide no relevant history of that lesion. I suggested a biopsy. The patient was concerned about healing near her bra strap and asked about other options. I described 2-GEP, and she was relieved at the possibility of skipping a surgical wound. The result came back LINC positive/PRAME negative, which indicated that the probability of melanoma assessed by dermatopathology was less than 10%. My policy is to biopsy all gene expression-positive lesions. With LINC positive/PRAME negative results, I generally tell patients that some early molecular changes are occurring in their lesion, but chances for melanoma are pretty low. However, in this case, the clinical picture was concerning enough that I did not provide this standard reassurance to the patient. She was biopsied 10 days after the test was performed. The result was melanoma in situ. The excision was performed 14 days later, and the patient visits our office for frequent skin checks.
This was an interesting case also from the perspective of it being the second melanoma in situ that I diagnosed on a LINC positive/PRAME negative result in 1 week. During my entire first year of integrating 2-GEP into my practice and using the test more than 300 times, none of my approximately 20 LINC positive/PRAME negative results led to a pathologic diagnosis of melanoma. I was beginning to be lulled into a false sense of security. My recent results were a major wakeup call. The bottom line is, ignore LINC positive/PRAME negative results at your and your patient’s peril.
Case 2: Double positive for LINC and PRAME genes
A 54-year-old man presented with concern about a changing mole on his left forearm. The patient had a family history of melanoma (father) but no personal history of melanoma. A full exam revealed signs of chronic sun damage and a concerning 6-mm darker and lighter brown asymmetric macule on the left forearm (Figure 2).
On dermoscopy, there was an irregular pigment network and no signs of reassuring horn cysts. I told the patient that the lesion needed to be sampled and discussed options. We decided on 2-GEP, which resulted in a double positive test result, with both LINC and PRAME gene expression consistent with gene expression seen in melanoma detected.
The patient returned 12 days later for a saucerization biopsy of the lesion. The biopsy showed a 0.5-mm melanoma with a tumor staging of T1a. A Castle 31 gene expression profile test was performed on the surgical biopsy specimen and revealed a low-risk result. With the low-risk Castle result and a T1a tumor staging, the patient’s 5-year survival rate is greater than 99% and his long-term prognosis is excellent. The melanoma was excised 19 days later. The patient will be closely followed in the clinic and was instructed to perform a monthly self-scan and wear sunscreen.
Ferris LK, et al. Dermatol Online J. 2019;25:1-8.
Ferris LK, et al. J Invest Dermatol. 2019;doi:10.1016/j.jid.2018.10.041.
Ferris LK, et al. Melanoma Res. 2018;doi:10.1097/CMR.0000000000000478.
Gerami P, et al. J Am Acad Dermatol. 2017;doi:10.1016/j.jaad.2016.07.038.
Hornberger J, et al. JAMA Dermatol. 2018;doi:10.1001/jamadermatol.2018.1764.
Rivers JK, et al. Skin Therapy Lett. 2018;23:1-4.
Disclosure: Brouha reports he is on the scientific advisory board for DermTech and a paid investigator in DermTech-sponsored clinical trials.