Faculty Bibliography

A diagnosis of late stage melanoma is associated with significant mortality. From a public health perspective, knowledge of geographic disparities in late-stage diagnoses can inform efforts to facilitate the diagnosis of earlier stage, highly curable melanomas. We conducted a county-level analysis of melanoma in New York State to identify communities that may benefit from pilot health interventions to reduce the burden of late stage melanoma. From 1995 to 2016, late-stage melanoma incidence increased from 1.5 to 2.8 cases/100,000 in NY State. We found statistically significant associations between decreased county-level health system access (including physician density and resident educational status), and increased county incidence and proportion of late-stage disease among diagnosed cases (p<0.001 for both). Increased county-level socioeconomic status (SES), (including measures of resident wealth and medical insurance status), was positively associated with greater late-stage incidence (p<0.001). However, decreased county-level SES was positively associated with a greater proportion of late-stage disease among cases at diagnosis (p=0.009). Counties with reduced access to physician services and lower SES may be suitable for pilot interventions promoting recognition and diagnosis of early stage melanomas to reduce late stage diagnoses and associated mortality.

Nevisense is an FDA-cleared device to aid in diagnosing melanoma. Using a non-invasive probe, the device measures electrical impedance spectroscopy (EIS) of target skin lesions. While EIS has demonstrated high sensitivity in diagnosing melanoma, its impact on a clinician's diagnostic confidence remains unknown. We conducted a pilot study evaluating whether the addition of EIS scores to clinical and dermoscopic images increases diagnostic confidence, accuracy, sensitivity, and specificity for students and dermatologists when evaluating lesions clinically suspicious for melanoma. Three pigmented lesions specialists and three 4^th year medical students completed an online survey to evaluate 34 melanocytic lesions suspicious for melanoma. For each lesion, participants provided their diagnosis, biopsy recommendation, and confidence in diagnosing a lesion as benign or malignant based on history and clinical and dermoscopic images, and again after receiving an EIS score. Addition of EIS scores increased mean biopsy sensitivity for melanoma/severe dysplastic nevi (DN) from 70% to 84% (p =.014) and mean diagnostic accuracy from 74% to 86% (p =.005). Mean diagnostic confidence increased for 29/34 lesions, of which 26 were accurately diagnosed by >=4 evaluators. Increases in diagnostic confidence were significant for common melanocytic nevi, DN, and melanoma, for both students and dermatologists (all p <.05). Use of EIS may increase clinicians' confidence to provide greater reassurance regarding dermoscopically equivocal lesions such as DN. EIS thus has the potential to help clinicians better alleviate patients' anxieties during skin exams. EIS may also improve management of melanocytic lesions suspicious for melanoma among novice and expert diagnosticians, though further investigation is needed to determine if these findings translate to clinical settings. NB: All authors except for Ms. Fried are team members for a separate study utilizing a Nevisense device, loaned to NYU by Scibase.
Copyright

BACKGROUND:Melanoma lacks validated blood-based biomarkers for monitoring and predicting treatment efficacy. Cell-free circulating tumour DNA (ctDNA) is a promising biomarker; however, various detection methods have been used, and, to date, no large studies have examined the association between serial changes in ctDNA and survival after BRAF, MEK, or BRAF plus MEK inhibitor therapy. We aimed to evaluate whether baseline ctDNA concentrations and kinetics could predict survival outcomes. METHODS:mutation-positive metastatic melanoma and brain metastases. Patients in cohort A of COMBI-MB had asymptomatic brain metastases, no previous local brain-directed therapy, and an ECOG performance status of 0 or 1. Biomarker analysis was a prespecified exploratory endpoint in both trials and performed in the intention-to-treat populations in COMBI-d and COMBI-MB. We investigated the association between mutant copy number (baseline or week 4 or zero conversion status) and efficacy endpoints (progression-free survival, overall survival, and best overall response). We used Cox models, Kaplan-Meier plots, and log-rank tests to explore the association of pretreatment ctDNA concentrations with progression-free survival and overall survival. The effect of additional prognostic variables such as lactate dehydrogenase was also investigated in addition to the mutant copy number. FINDINGS/RESULTS:mutation-positive ctDNA concentration was associated with worse overall survival outcome (hazard ratio [HR] 1·13 [95% CI 1·09-1·18], p<0·0001 by univariate analysis), independent of treatment group and baseline lactate dehydrogenase concentrations (1·08 [1·03-1·13], p=0·0020), in COMBI-d. A ctDNA cutoff point of 64 copies per mL of plasma stratified patients enrolled in COMBI-d as high risk or low risk with respect to survival outcomes (HR 1·74 [95% CI 1·37-2·21], p<0·0001 for progression-free survival; 2·23 [1·73-2·87], p<0·0001 for overall survival) and was validated in the COMBI-MB cohort (3·20 [1·39-7·34], p=0·0047 for progression-free survival; 2·94 [1·18-7·32], p=0·016 for overall survival). In COMBI-d, undetectable ctDNA at week 4 was significantly associated with extended progression-free and overall survival, particularly in patients with elevated lactate dehydrogenase concentrations (HR 1·99 [95% CI 1·08-3·64], p=0·027 for progression-free survival; 2·38 [1·24-4·54], p=0·0089 for overall survival). INTERPRETATION/CONCLUSIONS:-mutant ctDNA measurements could serve as independent, predictive biomarkers of clinical outcome with targeted therapy. FUNDING/BACKGROUND:Novartis.

Importance/UNASSIGNED:Dermoscopy education in US dermatology residency programs varies widely, and there is currently no existing expert consensus identifying what is most important for resident physicians to know. Objectives/UNASSIGNED:To identify consensus-based learning constructs representing an appropriate foundational proficiency in dermoscopic image interpretation for dermatology resident physicians, including dermoscopic diagnoses, associated features, and representative teaching images. Defining these foundational proficiency learning constructs will facilitate further skill development in dermoscopic image interpretation to help residents achieve clinical proficiency. Design, Setting, and Participants/UNASSIGNED:A 2-phase modified Delphi surveying technique was used to identify resident learning constructs in 3 sequential sets of surveys-diagnoses, features, and images. Expert panelists were recruited through an email distributed to the 32 members of the Pigmented Lesion Subcommittee of the Melanoma Prevention Working Group. Twenty-six (81%) opted to participate. Surveys were distributed using RedCAP software. Main Outcomes and Measures/UNASSIGNED:Consensus on diagnoses, associated dermoscopic features, and representative teaching images reflective of a foundational proficiency in dermoscopic image interpretation for US dermatology resident physicians. Results/UNASSIGNED:Twenty-six pigmented lesion and dermoscopy specialists completed 8 rounds of surveys, with 100% (26/26) response rate in all rounds. A final list of 32 diagnoses and 116 associated dermoscopic features was generated. Three hundred seventy-eight representative teaching images reached consensus with panelists. Conclusions and Relevance/UNASSIGNED:Consensus achieved in this modified Delphi process identified common dermoscopic diagnoses, associated features, and representative teaching images reflective of a foundational proficiency in dermoscopic image interpretation for dermatology residency training. This list of validated objectives provides a consensus-based foundation of key learning points in dermoscopy to help resident physicians achieve clinical proficiency in dermoscopic image interpretation.

The growth of molecular technologies analyzing skin cells and inherited genetic variations has the potential to address current gaps in both diagnostic accuracy and prognostication in melanoma patients or in individuals at risk for developing melanoma. In part II of this continuing medical education article, novel molecular technologies are reviewed. These have been developed as adjunct tools for melanoma management and include the Pigmented Lesion Assay (PLA), myPath Melanoma, and DecisionDx-Melanoma tests, and genetic testing in patients with a strong familial melanoma history. These tests are commercially available and marketed as ancillary tools for clinical decision-making, diagnosis, and prognosis. Here we review fundamental principles behind each test, discuss peer-reviewed literature assessing their performance, and highlight the utility and limitations of each assay. The goal of this article is to provide a comprehensive, evidence-based foundation for clinicians regarding management of patients with difficult pigmented lesions.

Managing the balance between accurately identifying early stage melanomas while avoiding obtaining biopsy specimens of benign lesions (ie, overbiopsy) is the major challenge of melanoma detection. Decision making can be especially difficult in patients with extensive atypical nevi. Recognizing that the primary screening modality for melanoma is subjective examination, studies have shown a tendency toward overbiopsy. Even low-risk routine surgical procedures are associated with morbidity, mounting health care costs, and patient anxiety. Recent advancements in noninvasive diagnostic modalities have helped improve diagnostic accuracy, especially when managing melanocytic lesions of uncertain diagnosis. Breakthroughs in artificial intelligence have also shown exciting potential in changing the landscape of melanoma detection. In the first article in this continuing medical education series, we review novel diagnostic technologies, such as automated 2- and 3-dimensional total body imaging with sequential digital dermoscopic imaging, reflectance confocal microscopy, and electrical impedance spectroscopy, and we explore the logistics and implications of potentially integrating artificial intelligence into existing melanoma management paradigms.