Introduction: Non-invasive adjuncts to visual assessment of pigmented lesions may reduce biopsies of benign lesions without compromising melanoma detection. A non-invasive genomic melanoma rule-out assay analyzes RNA extracted from stratum corneum cells for PRAME and LINC00518, two genes commonly expressed in melanomas but less often in benign lesions. This study sought to characterize performance of this test in a large patient cohort tested in the real-world clinical setting. Methods: The test was applied to suspicious pigmented skin lesions at 63 U.S. dermatology and primary care practices. Test results (positive / negative) were compared to pathology diagnoses (melanoma / not melanoma) for lesions that were biopsied and to follow-up visual examination for those that were monitored. Results: Of 19,653 total lesions evaluated, 17,858 (90.87%) tested negative. Biopsy results and / or follow-up examinations were available for 5,096 lesions, with median and mean follow-up duration of 352 and 341 days, respectively. For melanoma, sensitivity was 95.8% and specificity was 69.4%. Positive predictive value (PPV) was 13.4%, and NPV was 99.7%. For melanoma and ‘borderline’ lesions combined, sensitivity was 94.2%, specificity was 71.2%, PPV was 20.8%, and NPV was 99.3%. Conclusion: The results suggest this noninvasive test can facilitate distinction of melanoma from its benign simulators, increasing the proportion of pigmented lesions that can be safely managed with surveillance rather than biopsy and/or excision.

Skin cancer risk is increased by exposure to ultraviolet radiation (UVR). Because UVR exposure accumulates over time and lighter skin is more susceptible to UVR, age and skin tone are risk factors for skin cancer. However, measurements of somatic mutations in healthy-appearing skin have not been used to calculate skin cancer risk. In this study, we developed a noninvasive test that quantifies somatic mutations in healthy-appearing sun-exposed skin and applied it to a 1038-subject cohort. Somatic mutations were combined with other known skin cancer risk factors to train a model to calculate risk. The final model (DNA-Skin Cancer Assessment of Risk) was trained to predict personal history of skin cancer from age, family history, skin tone, and mutation count. The addition of mutation count significantly improved model performance (OR = 1.3, 95% confidence interval = 1.14-1.48; P = 5.3 × 10-6) and made a more significant contribution than skin tone. Calculations of skin cancer risk matched the known United States population prevalence, indicating that DNA-Skin Cancer Assessment of Risk was well-calibrated. In conclusion, somatic mutations in healthy-appearing sun-exposed skin increase skin cancer risk, and mutations capture risk information that is not accounted for by other risk factors. Clinical utility is supported by the noninvasive nature of skin sample collection through adhesive patches.

BACKGROUND:Numerous melanoma-specific dermoscopic features have been described in invasive melanomas, while fewer features are found in melanoma in situ (MIS) and atypical nevi (ATN). Consensus regarding which features are critical for the differentiation of MIS from ATN has not been reached. PURPOSE:Determine 1) whether there are dermoscopic features that differentiate early MIS from ATN, and 2) whether non-invasive assessment of genomic biomarkers (LINC00518 and PRAME) can aid in patient management. METHODS:From 2018 to 2023, 56 melanomas were evaluated for 5 clinical and 13 dermoscopic features and melanoma-associated genomic biomarkers. Two groups of ATN with positive and negative genomic biomarkers were randomly selected for comparison. RESULTS:All melanomas in this study expressed one or both melanoma-associated genomic markers. MIS had an average of 3.90 (range, 2-7) of the 13 dermoscopic features, while invasive melanomas had an average of 4.44 (range, 3-6). Sixteen of 40 (40%) MIS and 3 of 16 (18.8%) invasive melanomas had 3 or fewer dermoscopic features. These findings were comparable to those observed in both ATN groups. The most common dermoscopic features were absent or diminished pigment network, regression structures, and granularity. This combination of features was most helpful in identifying lesions for genomic testing. CONCLUSIONS:Clinical and dermoscopic features alone could not differentiate MIS from ATN. Non-invasive genomic testing helped differentiate lower from higher-risk lesions and aid in clinical management decisions. Genomic testing was particularly helpful in patients with large numbers of lesions with several being considered for biopsy based on clinical and dermoscopic examination. J Drugs Dermatol. 2024;23(9):717-723. doi:10.36849/JDD.8454.

Non-invasive assessment of pigmented skin lesions is gaining adoption by clinicians and provides a tool for enhanced evaluation of lesions suggestive of melanoma. One such test detects gene expression of LINC00518 and PRAME in samples of pigmented skin lesions collected non-invasively via adhesive patches that can rule out melanoma with a negative predictive value (NPV) of 99%. The purpose of the current study was to further characterize this test's performance in patients with Fitzpatrick skin types IV-VI. Performance metrics for samples from skin types I-III (n=4047) and IV-VI (n=125) were calculated and compared. Since the prevalence of melanoma is known to be lower in skin types IV-VI, lesions that test negative in these subjects would be expected to have a very low probability of being melanoma (false negative result). To estimate performance conservatively in this group, a 10% false negative "handicap" rate was therefore imputed. Among the 4047 Fitzpatrick I-III lesions, 66 were diagnosed as melanoma by histopathologic assessment. Test performance in this group was sensitivity 0.97, specificity 0.91, and NPV 0.9994. Among the 125 Fitzpatrick IV-VI subjects, all three melanomas diagnosed by histopathology had been correctly identified by positive rule-out test results as higher risk. The test performance was sensitivity 0.91, specificity 0.94, and NPV 0.9974. Performance of the test in skin types IV-VI is similar to that in types I-III in the cohorts evaluated. Sensitivity and specificity were 90% or higher in both groups. Most importantly, the NPV was also similar for both groups, at greater than 99%.

Background: ​ Pigmented lesion evaluation remains a challenging aspect of dermatology. The DermTech Melanoma Test (DMT) is a non-invasive gene-expression test designed to rule-out melanoma. It consists of the pigmented lesion assay, which detects RNA products of Long Intergenic Non-Coding RNA 00518 (LINC00518) and Preferentially Expressed Antigen in Melanoma (PRAME), and an add-on assay for DNA promoter mutations in telomerase reverse transcriptase (TERT). This registry study examines the concordance of PRAME detection by polymerase chain reaction (PCR) in samples obtained non-invasively prior to biopsy and PRAME detection by immunohistochemistry (IHC) on the same lesions after biopsy.​ Methods: ​ Between April 2021 and March 2022, multiple geographically diverse sites throughout the US submitted data to a registry to assess real-world use of the DMT. Approximately 8,000 clinically atypical lesions were tested. After receiving the test result, providers followed their clinical judgement for biopsy decision. When lesions expressed genomic markers (LINC, PRAME, and/or TERT) and were biopsied, pathology reports were also submitted to the registry. The presence or absence of PRAME by immunohistochemistry (IHC) was reviewed and compared to the detection of PRAME by PCR from the DMT on the same lesion.​ Results: ​ At the 1-year mark of the registry, there were roughly 8,000 unique entries. Of those, 1,021 (12.8%) were positive for one or more of the DMT genomic markers. One thousand three lesions (98.2%) had records available. Pathologists used PRAME IHC for 102 lesions (10.2%). Of those, 40 (39.2%) were positive by IHC, and 62 (60.8%) were negative by IHC. PRAME positivity by PCR correlated with PRAME positivity by IHC in 35 of 40 lesions (87.5%). Conversely, PRAME was detected using PCR in 28 of 62 lesions (45.2%) where it was not detected using IHC. ​ Conclusions: ​ The higher sensitivity of PCR compared to IHC may explain the higher concordance when PRAME is positive by IHC than when it is negative by IHC. In this data set, when PRAME is positive by IHC it is usually also positive by PCR. When PRAME is negative by IHC, it can still be detected by PCR in a substantial percentage of cases. The increased sensitivity of PCR is likely due to several factors, including its detection of the PRAME mRNA and sampling of the entire lesion. As such, PRAME PCR status may aid pathologists in understanding the risk of melanoma even when IHC is negative. Further research is warranted to understand the clinical implications of PRAME PCR versus IHC positivity. ​

Supplemental Figure 1 shows the distribution of gene expression scores in the overall cohort.

Objective: Fitzpatrick Phototypes (FP) and sex are risk factors considered when assessing lesions for keratinocyte skin cancer (SC). National statistics for the US population report that SC is most frequently first diagnosed at the age of 65.

Supplemental Figure 6 shows a lesion from the face of a 73 year old male was classified as a desmoplastic melanoma, pure type (Breslow thickness at least 1.2 mm, mitotic index 0/mm2, ulceration absent). The score was -2.6 (false negative). Metastases to lymph nodes and lungs were detected 18 months after initial diagnosis.

Background: Non-melanoma skin cancers (NMSCs), including basal cell carcinoma (BCC) and squamous cell carcinoma (SCC), are responsible for more than 15,000 deaths each year in the US [1]. Current diagnosis of NMSCs relies primarily on biopsy and histopathologic examination. This study aimed to develop a non-invasive test for detecting BCC, distinguishing it from both SCC and non-cancerous skin diseases (NCSDs), using stratum corneum cells collected with adhesive patches. Biomarker discovery was performed in three steps to identify the most reliable and generalizable biomarkers. First, whole transcriptome RNA sequencing of 354 patients (94 BCC, 87 SCC, and 173 NCSD) identified 161 significantly differentially expressed genes in BCC versus SCC and NCSD. Second, targeted RNA AmpliSeq analysis of these 161 genes was performed on a set of 366 patients (108 BCC, 89 SCC, and 169 NCSD), resulting in 46 significantly (fold-change > 2; adjusted p-value < 0.05) differentially expressed genes, 30 genes with increased expression in NMSC, and 16 genes with lower expression in NMSC. Last, the top 37 genes were selected for model training and validation using qPCR. The top five biomarkers (TAGLN, FDCSP, LINC02167, FOXI3, and CASC15) individually showed Area Under the ROC Curve (AUC-ROC) of above 0.8. A machine learning algorithm, trained and tested with 10-fold cross-validation using the top 10 biomarkers was able to differentiate BCC from SCC and NCSD with AUC-ROC above 0.85. The results are promising, and further analysis and validation is currently underway.

Background: Cutaneous T cell lymphoma (CTCL) is a complex and heterogenous disease characterized by cutaneous infiltrates of aberrant monocolonal T-lymphocytes. Clinically and histopathologically, the most common form of CTCL (mycosis fungoides) often mimics psoriasis, eczematous dermatitis, and other common dermatoses. Diagnosis of CTCL is often delayed by 4-6 years from the time lesions initially appear, and typically requires multiple skin biopsies and blood tests.

Supplemental Figure 4 shows a lesion from the posterior thigh of a 70 year old female that was classified as a nodular melanoma, nevoid type (Breslow thickness 1.25 mm, mitotic index 2/mm2, ulceration absent). The score was +6.6. Brain metastases developed five years after initial diagnosis.

Supplemental Figure 3 shows a lesion from the forehead of a 57 year old male that was classified as lentigo maligna melanoma (Breslow thickness 1.0 mm, mitotic index 0/mm2, ulceration absent). The lesion was completely excised. The score was +4.0. Approximately 7 years after initial diagnosis, melanoma was detected within two of twenty lymph nodes from the right neck.

Background: The pigmented lesion assay (PLA) evaluates lesions suspicious for melanoma by analyzing skin tissue for preferentially expressed antigen in melanoma (PRAME) and long intergenic non-protein coding RNA 00518 (LINC) via RT-PCR. The majority of samples tested are from patients with Fitzpatrick Phototypes (FP) I-IV, where concerning pigmented lesions are common. However, suspicious pigmented lesions also occur in patients with FP IV-VI, the purpose of this study was to characterize assay performance in this population.

Background: Fitzpatrick Skin Phototypes (FSP) is a classification system used by clinicians to assess an individual's skin type into one of six grade levels based on the skin's melanin content and response to ultraviolet radiation. The FSP assessment can help identify risk factors for photodamage and help determine preventative treatment strategies. This study evaluated the difference in FSP perception between subject self-reporting and clinician-reported of FSP.

Supplemental Figure 5 shows a lesion from the back of a 43 year old female was classified as a superficial spreading melanoma arising within a pre-existing dysplastic nevus (Breslow thickness 1.6 mm, mitotic index 0/mm2, ulceration absent). The lesion was completely excised. The score was +4.6. Metastases to the soft tissue and skin of the upper chest were detected 56 months after diagnosis.

Background: Pigmented lesion analysis remains a challenging aspect of dermatology. The DermTech Melanoma Test (‘the test’) is a non-invasive genomic test designed to rule-out melanoma. It consists of the pigmented lesion assay, which detects RNA products of Long Intergenic Non-Coding RNA 00518 (LINC00518) and Preferentially Expressed Antigen in Melanoma (PRAME), and an add-on assay for DNA promoter mutations in telomerase reverse transcriptase (TERT). In previous studies, the test was found to have a negative predictive value ≥99%. This interim analysis of a registry study examines the genomic patterns of the Lentigo Maligna (LM) subtype of melanoma. Methods: Between April 2021 and March 2022, multiple geographically diverse sites throughout the US submitted data to a registry to assess real-world use of the test. Approximately 8,000 clinically atypical lesions were tested. After receiving the test result, providers followed their clinical judgement for biopsy decision. Histopathologic diagnoses for biopsied lesions were correlated with test results and all melanomas were sorted into LM subtype vs non-LM subtype. In addition, lesions that were noted to be on sun exposed skin and/or noted to have solar elastosis and called atypical melanocytic hyperplasia were included. Results: At the 1-year mark of the registry, there were roughly 8000 unique entries. Of those, 1003 expressed one or more genomic markers from the DMT and had records available, and 134 (13.2%) were found to be melanoma or melanoma in-situ. More than a third (n=46, 34.3%) of the melanomas were of the Lentigo Maligna sub-type, with 7 of those being Lentigo Maligna Melanoma (LMM). Seven additional lesions were called atypical junctional melanocytic hyperplasia (AJMH) on sun-damaged skin. This group of 53 LM, LMM, and AJMH lesions were all evaluated for correlations to the DMT markers. LINC was the most commonly expressed genomic marker (n=45, 84.9%), with PRAME (n=36, 67.9%) and TERT (n=24, 45.3%) following. Most invasive tumors (LMM) expressed all three markers (n=4) and all 7 expressed LINC. Conclusion: While the original validation study included the LM subtype, this interim registry analysis demonstrates real-world use of the DMT in assessing pigmented macules concerning for LM. Over 1/3 of the melanomas reported in the registry were LM subtypes. LINC had a higher correlation with the lentigo maligna subtype of melanoma and was present in all invasive tumors. While AJMH is considered borderline, it may be worthwhile in the clinical context to group AJMH lesions with LM due to similarities in treatment.

Supplemental Figure 7 shows a lesion from the posterior shoulder of a 54 year old male was classified as compound dysplastic nevus with severe atypia. Melanoma could not be excluded on histopathologic examination. The score was +4.1. This case was considered a false positive because metastases were not detected during a follow-up period of seven years.

Supplemental Figure 2 shows a lesion from the right upper back of a 71 year old male that was classified as a superficial spreading melanoma ex nevus (Breslow thickness 0.69 mm, mitotic index 1/mm2,ulceration absent). The score was +4.1 Metastases to the soft tissue of the right neck developed 16 months after initial diagnosis.